John de Moncaux, and William Porter, on numerous occasions. John. Habraken's concept of typology, ...... it; and a nest is built by birds. John Ruskin. In today's ...
after amnesia learning from the islamic mediterranean urban fabric
Attilio Petruccioli
after amnesia learning from the islamic mediterranean urban fabric
icar
Dipartimento di Ingegneria Civile e Architettura Politecnico di Bari Via Orabona 4, 70125, Italy © Attilio Petruccioli settembre 2007
Progetto grafico e impaginazione Nino Perrone Composizione tipografica in Bembo Standard by Stanley Morison, Monotype Staff Scala Sans by Martin Majoor Stampa Grafica & Stampa, Altamura (Bari) - Italy ISBN 88-95006-03-8 978-88-95006-03-1
contents
9
introduction
15 15 20 23 25 27 28 34 38 48 60
the fourth typology Building and Civilization Structure and Type Type and the Modern Standard The A-priori Type The A-posteriori Type Interpreting Serial and Organic Structures The Typological Process Interpreting and Design Notes
65 67 73 85 85 91 95 104 117 125
the building type The Elementary Type The Courtyard House Reading the Courtyard as an Organism of Systems of Structures of Elements: Fez Mostar Aleppo Mediterranean Typological Processes The Colonial Tenement House in Algiers Notes
131 131 136 138 153 162 170 174
the building tissue Aggregative Principles On the Urban Fabric Routes and Building Tissues The Courtyard Tissue Interpreting Building Tissue and Urban Fabric Cul-de-sac Notes
179 181 192 205 215 223 232
the urban organism The Urban Nucleus Type The Labyrinth of the Hierarchies Interpreting the Urban Fabric One City does not Hide Another The City: Neither Spontaneous Nor Created Notes
this publication was made possible thanks to a generous grant from the foundation max van berchem, geneva. The Max Van Berchem Foundation is a scientific foundation established in Geneva, Switzerland, in memory of Max Van Berchem (1863-1921), the founder of Arabic epigraphy. Its aim is to promote the study of Islamic and Arabic archeology, history, geography, art, epigraphy, religion and literature.
This book is dedicated to Alessandro Giannini
9
introduction
Architects easily forget that one must follow the style of the place rather than that of the time. Gunnar Asplund There are two main categories of architects. To the first belong the few fortunate people endowed with a natural ability to understand and manipulate space; only a hundred or so of those are born in any century. To the second belong all the others, including myself, and it is to them that this book is devoted. I am convinced that the art of designing can be learned. Architecture is a beautiful and exciting game but it should be carried out with the discipline of a solid technique upon which it is then possible to graft a measure of fantasy and even the exceptional. Without discipline, architecture is reduced to painting or sculpture. That is to say, abandoned to the temptations of novelty for novelty’s sake, the systematic infringement of the already infringed, an acrobatics of design reduced to the claims of personal style. Today many architects simply seek fame and wealth. But differently from the past, this comes at a time when architects have lost their grip on the city. Others have the more commendable goal of correctly constructing valuable buildings that strive for harmony with their context. Still others have the illegitimate unrealistic hope of formulating a grammar of architecture that can solve, once and for all, the problem of making architecture. This can go so far as to propose a universal manual capable of solving the problems that impact lack of communication between cultures. In the past such attempts based on personal criteria and self-referential methods have had ephemeral success with results that have invariably inflicted serious damage on the built landscape. Fortunately, this book is neither a treatise nor a grammar. It is merely an attempt at suggesting a route that can lead to a grammar that already exists and is preserved in the built landscape. It is a method for rigorously interpreting a fabric, a city, and a territory. Today, when people are knowingly destroying the environment and erasing the past, it may seem like a contradiction to maintain that every human action has left deep traces in the territory. I am not simply referring to the macroscopic phenomenon of the Roman centuriatio, but also to the primitive routes on top of ridges, landmarks in the urban topography, and signs on facades that signify permanence. These are the elements of spatial and historical continuity that fully represent a record of civilization. Edmond Desmolins said: “History belongs to the territory and the awareness that man has of it.” He suggests that the physical form of a territory writes the pages of history. Furthermore, he recognizes an important subjective component of its reading, a deep level of consciousness. The built landscape around us is an expression of our civilization, and monuments are symbols of civic, moral, and religious
10
after amnesia
values. The building intended as dwelling represents our deepest level of consciousness. The monument and its form have always been an object of semantic manipulation and cultural transfer. We cannot explain Filarete’s treatise on Sforzinda or the slender towers of Urbino’s Ducal Palace without referring to the Indian world or Ottoman Turkey. This is because the monument is transferred by modes, which, as Quatremere de Quincy has pointed out, can be, and often are, copies. In contrast, residential buildings resist transfer. They are rooted in a place because every civilization is in flux and continuously readapts its house to both new needs and requirements. The urban fabric is the main concern of this work, both as a source for the principles of design and as a way of making our analysis more sophisticated. Unfortunately, the innumerable studies on vernacular building are not useful for our purpose because they do not go beyond a simple description of the object. The deductive process is deceptive because it considers only the facts as related to the method. The method suggested in this book is inductive. It is based on those principles intrinsic to the built landscape that are ordered by the concept of type. It is not a question here of reproducing taxonomic criteria, or of a functional mechanistic vision that reduces type to standard, or even a formalistic vision that reduces type to a schema extrapolated from history. It is, on the contrary, an idea of type intimately connected to social body that presides over the examples in the built landscape through history. Type is history. Like that of an artisan, this idea of type belongs ultimately to spontaneous consciousness. It is the idyllic condition that no longer exists. But it is always possible to rediscover it through a critical reconstruction of the available fragments living in the principles preserved in the fabric. Unlike my predecessors, I admit that if type is history it is an essential component of space. A classificatory method allows us to establish a preliminary generic order that we analyze in reality. But it is only the concept of process applied to type that permits us to penetrate the mechanisms of mutations of the city and the territory. The idea of applying typological process to the reading of the city is the major contribution of this book. Normally the reader expects the sequence of analysis to follow from the largest to the smallest scale of study. In other words, one proceeds from the level of a territory, which contains all the other elements of measure, itself being the most complex, then progresses downward to the city and the house. We assume that this is always the synthetic way we look at reality. However, I have chosen to move analytically from the smallest object, that which is most familiar in its structure to the architect and most intimate to the human scale. In this manner the analysis moves through progressively larger contexts. In sum, I believe that a typological process moving from the global to the minute loses its vigour of perspective. Objects remain some sort of sum total negating the accumulation of complexities that would otherwise turn up throughout multiple levels. I do not want to create the impression that this manner of approach and the numerous drawings in the text represent more than an introduction to the problem of understanding the built landscape. This
introduction work is not intended as a substitute for the experience of studying the real fabric. Only continual observation and practice sharpen the eye. It is also necessary to be aware of the limits of the method, since a method for obtaining knowledge of reality should not be more complex than the reality itself. What is proposed here is a project that helps facilitate an understanding of the design process. The region chosen for study is the Islamic Mediterranean basin, the long geographical strip that runs from Spain to Bosnia via the Maghreb and the Levant. This choice was determined by two considerations. First, this is the geographical zone where I have thirty-five years of experience related to urban fabric. Second, the Mediterranean has had its own cultural koinË for many centuries, to which all cultures along it shores have conformed. Islam is for me the true heir of the classical world that created that koinË. The classical patrimony, which was disrupted in the West, was absorbed and elaborated into a new synthesis in the East. The Arab city is also very useful for understanding our cities. The result is that pre-existent structure is contained in a structure of successions. And although this does not readily reveal its previous mutations, it is always there. It is my opinion that the ultimate achievement of the architect is to design an architecture that seems to have always existed. I decided not to use photographs in this book. While photography has the merits of accelerating the documentation of architecture and introducing it to the public, it also has the tendency to make students and architects analytically lazy. Whereas, on the one hand, drawings are not restricted to the objects framed by the camera’s eye, they do enhance our perspective and understanding of the object studied through the rendering process itself. In this way drawings may contribute to greater analytical clarity and objectivity. Drawings in the book represent a parallel route to the reading. They organize the points made in the analysis. The importance of the measured architectural survey for penetrating the mysteries of the buildings and the secrets of the urban fabric should not be underestimated. For this reason I continue to take groups of students to sites around the world on enthusiastic campaigns of urban and architectural documentation. The sources for every drawing are cited with acknowledgment of the amount of time the graphic work required. Drawings and architectural surveys produced in my courses will find the students’ names acknowledged. Most of the drawings borrowed for this book were done by me or a draftsman under my supervision. These were often done from serial sketches subject to continual corrections and revisions. These drawings have no credits. In closing, before delivering a manuscript to the printer, it is the author’s duty to thank those who contributed ideas, suggestions, and discussion to the work. First of all, I have had profitable exchanges of opinions with my colleagues at MIT, Stanford Anderson, Julian Beinart, John de Moncaux, and William Porter, on numerous occasions. John Habraken’s concept of typology, as collective social product, influenced my thought. Claudio D’Amato Guerrieri, with whom I shared the early years of architecture school and the past few years directing the School of
11
12
after amnesia
Architecture of the Bari Polytechnic, has discussed this book with me day after day. I am particularly indebted to Giancarlo Cataldi, who twenty-five years ago responded to my uncertainties and frustrations in trying to devise a method based on the experimental work of Saverio Muratori and Alessandro Giannini, who introduced me to the secrets of territorial studies. Several friends read and gave suggestions; among them I want to mention Mauro Bertagnin, Renata Holod and Anne Vernez Moudon. In addition, I would like to express gratitude to the new generations well represented by my teaching assistants Annalinda Neglia, Calogero Montalbano, Giuseppe Rociola, Claudio Rubini and Domenico Catania, my companions on excursions to Islamic countries. With the diligent and irreplaceable help of Rupinder Singh this book, originally written in Italian, was patiently rephrased and revised into English. Ellen Shapiro, guardian angel of all my writings, edited the text with great professionalism and good humour. I owe gratitude to Shakeel Hossain and Giuseppe Mazzone, who lent their magical computer skills to the textual and graphic layout. Attilio Petruccioli
A sketch showing a chart for the classification of the level of social and physical complexity. A work in progress.
15
the fourth typology
That which distinguishes truly original minds is not being the first to see something new, but seeing as new something old, well-known, seen and ignored by all. Friedrich Nietzsche Human All Too Human The cornerstone for my method of analyzing the built environment is a concept of type that is different from others. According to this definition, type is not a manipulation of morphology or banal classification of functions, but a universal concept manifested in built forms that are rooted in the historical process and social behaviour. This concept of type has two manifestations that are two sides of the same coin: when society is in a balanced state in its evolution it shows one side, and when it finds itself in a moment of structural crisis it shows the other. The first task is to separate into more readable parts the complexities of the built world and then to discover the origins of the components of the typological process, because in these origins or their approximation lie the explanations for complex phenomena. The reason for this division into readable parts and effort to uncover origins is my conviction that any instrument created to understand reality should not be more complex than the reality itself.
Fig. 1 Mestre, Italy. The house of a sea captain who retired and brought his ship home: A personal solution to dwelling
building and civilization A house is essentially a communal production: one man cannot build one house, but a hundred men can easily build a hundred houses. Hasan Fathy The long wave of globalization by its very nature tends to transform built reality into a kaleidoscope of images. Through specialized journals and mass media, using publicity techniques, it always presents everyday architecture as an extraordinary event. The repeated technique leads to a boring panorama of images, the exceptionality of the banal. Everyday architecture, building, has nothing banal about it. On the contrary, its “ordinariness” is its character, which mirrors the life of common people, who have always built their environment with it. Precisely because building is the most direct expression of a civilization, on a par with language and cuisine, it is the most directly affected by the current crisis in the architectural discipline. After the tabula rasa brought about by the Deconstructivist and Poststructuralist movements, arising
Fig. 2 The Îlot Candie sport centre designed by Massimiliano Fuksas. The facade is reduced to mere scenography, demonstrating the power of the architect and the critical indifference of the user. Fuksas refers here is the slightly inclined house of the Sacro Bosco at Bomarzo, a famous Renaissance infringement of building logic. But the suburban landscape in which the centre is located is incoherent and deprived of structure, and the provocation has little effect.
16
after amnesia
from the premises of the antihistorical attitude of the Modern Movement, any hypothesis of refoundation of the body of the discipline must pass through the rediscovery of building and its ethical and civic values. One task remains, and that is to clarify what the term “building” means and inserting into this term the role of history, culture, time and evolution. Nikolaus Pevsner’s observation that “a bicycle shed is a building; the Licoln Cathedral is architecture” seems to say that almost any structure that defines a space sufficient to move about in is a building, but architecture applies only to those buildings conceived with an aesthetic effect in mind.1 Pevsner’s extreme definition stems from a traditional art historian’s bias towards aesthetic interpretation and concern with defining the realm of the monument. In his last book, A History of Building Types,2 Pevsner does not go much beyond the nineteenth-century functional types (the railway station, hospital, prison, etc.), and he treats building types as monuments, if they fall within his definition of architecture, i.e., represent new or exceptional phenomena. To develop my definition, I will resort to etymology. Webster defines building as follows:3“1. A roofed and walled structure built for permanent use (as for a dwelling); 2.The art or business of assembling material into a structure.” The first definition embraces not only the bicycle shed but also the idea of an architectural organism; the second, the activities such as roofing and walling that are the elementary tectonic operations used in enclosing a space. Underlying this idea of an architectural organism is the dwelling, that steadfast structure of any social arrangement which generically corresponds to the minimum unit of the nuclear family. This definition of building then introduces the concept of structure, which is made explicit in the second, “the art or business of assembling material into a structure.” The “art of assembling” clearly refers to skill or experience in building. Building is all these things and more. For me, it is the art of constructing an edifice in an appropriate and correct manner, while at the same time being aware of the greater context provided by the spatial and historic continuum of houses, the neighborhood, the city, and the surrounding area.4 If a distinction between building and architecture must be made, it should be a dialectical distinction and not one set by Pevsner’s terms. For in addition to Pevsner’s buildings that aspire to the dignity of academic publication are the other more numerous ones that simply aspire to dignity. Building has always been ingrained in man’s collective and individual conscience to the extent that it stands second only to language as the means of cultural expression. From the social perspective, a difference between building and architecture did exist in the past, but it operated in the realm of civic conscience: people commissioned architects only for monuments. The origin and development of special buildings, or monuments, determined and formed the role of the architect. There was, however, no difference in the technical know-how between building and architecture. The distinction between building and architecture should thus be viewed as a critical judgment. A building is architecture when it radiates a perfection of expression that gives it the aesthetic value of a work of art. However, to qualify as art both architectural works and monuments
the fourth typology must have a historical-civic meaning or they would be abstract and individualistic in the context of the period in which they were conceived. The work of the great German master, H. Tessenow, is distinguished by this sense of artistry coupled with a moral attitude. As Karl Scheffler states in the introduction to the German edition of Tessenow’s Die Architektur der Groszstadt, “in architecture - as in life itself - individuality and originality are not as important as an overall effectiveness and impact that allow the individual elements of a work to lose their significance.”5 Although Tessenow uses the word “architecture,” his commentary is more about building than architecture, according to the categories that will be used here. The emphasis here will be on building rather than architecture, ordinary structures rather than masterpieces, the urban continuum rather than the outgrowths, and the continuity of enduring building processes which distinguishes architecture of great civilizations rather than the single, extraordinary event. It is this frame of reference that determines the tone of the built environment, described by Ludovico Quaroni as the “concrete and unexceptional spaces that are still of primary social importance, requiring a dignified and civil structural ‘order’ regardless of imposed limits, the quality of which must be upheld and protected especially in relation to the fact that it defines the living space for all”.6 Quaroni’s description further extends our concern and forces us to look holistically at the built environment. The Italian historian Gustavo Giovannoni7 was perhaps the first to identify the urban fabric as a connective tissue of houses and discussed its ethical and civic value. For him the cultural and civic significance of building can be understood only in the context of urban history. Without an intimate understanding of the urban tissue, urban history is technical, an abstract account of zoning and the development of road networks and public spaces. It is the picture of society in which the social dimension of the citizen is missing. A history of building that lacks a greater physical and social dimension is nothing more than a traditional architectural historical account of so-called vernacular architecture. If we carefully examine the elusive structure of such spontaneous Mediterranean cities as Cairo, Capri,Venice, and Tunis, where harmonious order is interwoven with apparent disorder, what results is a set of unique irreducible characteristics, where the personality of every neighbourhood is distinct, but at the same time is part of the polyphonic chorus of the place. Cairo, Capri,Venice, and Tunis each reveal the artistry of the urban fabric and force us to replace our romantic preconceptions about works of art as abstract and personal with the idea that works of art are spontaneous social interventions. In this way, it is possible to view society as able to create art even in everyday things. The ideals and history of any given cultural-historical region are expressed in the language used by the common people. The language may be enriched by its great writers, but they find their matrix in that language. Amsterdam is an example of such a harmonious balance. By the sixteenth century Amsterdam had become home to a stable, enlightened, and industrious middle class. When in 1607 the city needed to expand beyond its medieval perimeter, it happened in an almost biological continuum. Three concentric canals modelled on the existing canal were
17
18
after amnesia
thoughtfully integrated into the urban fabric, and all new building was buffered within them. What is extraordinary is that constructing this extension of Amsterdam required only the agreement of its citizens. The archives hold not a single design or planning project. A homogeneous bourgeois society spontaneously produced a blueprint for the town plan, which has since been rightfully considered an urban monument. In the second half of the nineteenth century, building as a product of spontaneous consciousness faced an unprecedented crisis and eventually its natural relation with the culture was severed. Rapid economic growth fueled by industrialization, as well as the specialization of the building industry itself, took by surprise those who venerated the existing “old” city. In addition, population growth caused the unprecedented outward expansion of those European cities just emerging from a state of semistagnation. The pre-industrial expansion of these cities had been inward and was achieved either through the erosion of public space or the deliberate transformation of existing structures. The new demographic and economic vitality combined with new mechanical means created new demands and functions. To meet them, traditional builders had only their limited experience and expertise derived from earlier changes in the local urban fabric and building practices and thus were only partially successful. The tastes of these medieval builders were formed by a common background and training, and their relation to building itself reflected this formation. The new bourgeois culture, in contrast, began to conceptualize buildings intellectually, as one can see in the universal forms of neoclassical architecture. Their response was primarily quantitative, leading to the reduction of a project to a simple question of style, in contrast to the spontaneous richness of the medieval town. Building became the monopoly of the architect and the owner: “building one’s house” became “finding one’s house on the market.” Since that time the fact of working on special buildings has moulded the professional figure of the architect. In addition to the inadequacies of the builders, nineteenth-century urbanism was adversely affected by the incapacity of the bureaucratic culture to assimilate the pre-industrial, medieval city. The Enlightenment, concerned with the abstract behaviour of people, eventually conceived the city in terms of mechanical problems like transportation and sanitation, refusing to understand the existing city beyond the superficial image of the medieval crowded narrow winding streets. The intellectual position of the Enlightenment also proved inadequate for facing the challenge of the new era; so did Romanticism in its vain desire to favour morality and emotion, and Positivism because of the narrowness of its vision of reality. The Enlightenment inspired countless modern interventions in historical cities, from the nineteenth-century attempts to “liberate” monuments (e.g., Hausmann’s gutting of Paris) to Le Corbusier’s famous Plan Voisin. I do not want to deny the dignity of nineteenth-century construction, urbanism, and architecture, but as a result of its conflictual relationship with the past it betrayed an exhaustion of ethical and expressive impulses.8 Our focus here is the Islamic Mediterranean, where the Cartesian “thinking being” produced the “technological society”, as Heidegger calls it, and did not plant any roots and, we can safely add, that building,
the fourth typology culture, and history were never separated. For example, several cities in the Middle East such as Damascus, and especially Aleppo, and Fatimid Cairo experienced almost uninterrupted growth from a classical period onwards. As a result of their commercial importance and political integration in the Ottoman Empire, from the sixteenth century onwards the populations of the Middle Eastern cities grew steadily.9 The urban fabric continued to expand, primarily according to building processes emphasizing coherent modification and complying with custom. In the second half of the nineteenth century, these cities (including Alexandria after 1830) grew considerably as a result of their opening up to Europe, as they adopted the imported cultural models that symbolized progress. New zones and infrastructure were randomly built and principally benefited a very limited social class. Despite both the imperial imagery of imported European architectural forms and an alien technology, it is important to remember that local forms and processes continued to manifest themselves and at times gave rise to interesting hybrids10 in these colonial cities. The fundamental mistakes of architects in the XXth century have always arisen from an artificial separation of the functional and formal attitudes. In fact, as Vitruvius understood two thousand years ago, owing to the intrinsic unity of the architectural organism, the three components of distribution, structure, and form cannot be separated. These two opposed attitudes when applied to design have created irreparable damage. They have generated either mechanical assemblages, Lego style, or abstract formal schemas that can be recycled in any context. Both mistakes are generated by the separation of typology from history. In the following paragraph we shall see that typology is history, since types are the projection of the society in time. An essential premise to a different concept of type that does not privilege the exception of the monument, but of the everyday fabric, is the re-evaluation of the notion of building: either as traditional technical expertise that has disappeared or as dwelling as minor architecture. Technical expertise has disappeared, not because it is obsolete, but because it has been driven out of the market by new materials. We have to articulate the idea of “diffused quality,” that is of the general use of correct construction and the coherent use of forms. This diffused quality was always present in the past, and it can still be recovered if we are willing to repair the rent in the fabric made by the Modern Movement and reestablish a continuity in historical processes. For this the architect will have to design buildings that are an expression of the society in which he lives and not of his own superego.
19
20
after amnesia
structure and type Socrates: I was almost running as I pondered this: a leafy tree is a product of nature; it is a building whose parts are the leaves, the branches, the trunk and the roots.While each of the parts suggests a certain complexity, I say that the whole of the tree is more complex than each of its parts. Paul Valery, Eupalinos ou l’architecte For a definition of type as a dynamic concept, structuralism and its application to linguistics can provide useful parameters. Piaget’s definition of structure might also be useful for the discussion.11 He says that “the structure, as opposed to the properties, of elements is a system of change that (because it is a system) involves rules. It is preserved and enriched by the process of change, but these changes never go beyond the boundaries of the structure or involve external elements. A structure has three characteristics: totality, transformation and self-regulation.” This definition, apart from its implied concept of totality, contains two important ideas: First, a structure is not immutabile. It is not set upon timeless foundations as logical-mathematical structures are. It undergoes constant change and can incorporate new elements that will enhance it. Second, to maintain the structure, these changes cannot go beyond its boundaries. This means that a self-regulatory mechanism exists that guarantees the maintenance of the internal laws and properties that define the structure. In other words, this is also the essence of the concept: When Louis Kahn asks the famous question to the brick. In effect, the brick (or the house) wants to be itself and cannot go beyond the limits imposed by its identity.12 The similarities between the characteristics of structure and of type are unmistakable, and many analogies can be drawn between the notion of the totality of a structure and the idea of organism in type. Piaget’s idea of a structure’s self-regulation suggests a rough definition of type as the impossibility of overstepping the limits imposed by its own internal structure. In other words, changes can generate only those elements that belong to the type and preserve its laws, otherwise the type simply mutates into something else. Following a crisis, a new building type can be generated from the preceding one without affecting the organic equilibrium of the distributive, static, and visual components that is, the Vitruvian triad of utilitas, firmitas, and venustas (fig. 5). In other words, a transition from one type to another can take place only if all of the Vitruvian components are affected and not just one or two of them.13 The evolution of the residential buildings on the Yemenite plateau illustrates this axiom. The houses of Sana’a are not simply tall, they are veritable towers. A greater emphasis on venustas in more recent types has resulted in a heavy and closed base supporting a lighter and perforated wall, and finally a mafrej, or rooftop room featuring one window-lined wall that accentuates the legibility of the parts. Firmitas was affected when the support structure was substantially changed from a box system using
the fourth typology supporting walls to a system similar to that used in modern skyscrapers, where the vertical pressures are generally concentrated on the central stairwell. This led to major changes in utilitas as the new distribution of rooms caused difficulty in separating off the harem, which ended up being relegated to a passing-through area. The result was a change in residential building type. Furthermore, the distinction that F. De Saussure makes between the terms langue and parole is particularly relevant.14 Langue refers to the totality of rules codified by the speakers of a language over time; violating them means losing the ability to communicate. Parole, on the other hand, refers to the individual’s idiosyncratic use of the langue code and its translation into discourse. Langue and parole are not contradictory; they are united by a dialectical process: one does not exist without the other. Children begin to speak as users of parole and not langue. The proper speech that results when langue and parole are combined is the product of an individual’s processing of the parole. The same criteria can be applied to the culinary arts, where a common body of recipes collected over time in a particular geographical area corresponds to langue, and the cook’s inspiration at a given moment to parole. The recipes may follow diets or trends, but in order to qualify as langue their variations must be dictated by rules. Over time, of course, each person’s way of preparing a favourite dish, or for that matter of dressing, speaking, or building a house will lead to the modification of that collective code. The idea of type similarly does not refer merely to a single set of codified rules, but carries with it all the limitations and contradictions derived from the particular use of that type, that is, the parole of the type. A type may be a part of a shared heritage, but the way it is used depends on the individual’s grasp of it. The built object is an intermediary between those ideas correlated to type and their intersection with personal histories. Everyone has an individual way of using words in a given context that often stretches the common codified use and will eventually change it. In much the same way over time the langue of architecture is bound to change. Langue allows for small-scale experimentation in much the same way as Sanskrit evolved into the various Prakrits and then united to form the various modern Indian languages, or the way classical languages met and evolved into the modern Western languages. Trial and error often determine change: in this sense we can say that modern languages are the fruit of the errors made by our ancestors. L. Hjemslev developed De Saussure’s dual concept and postulated the need for an entity halfway between langue and parole to represent the individual idiom that leads to change. L. Hjemslev divided language into three parts: the scheme or De Saussurian langue; the rules or norms; and usage, or the sum of a society’s customary way of speaking.15 Without overly forcing the comparison of building to language, we might say that our type is not so different from Hjemslev’s usage, and parole is comparable to the historically and geographically specific expression of a single building. Buildings, cities and territory are, of course, not written texts, and it was the habit of treating them as such in the sixties that led to the abuses of semiotic categories and to attempts to apply semiotics to architecture that succeeded only in confusing the issue. Communication
21
Fig. 3 The Yemeni house. Plans at the various levels. To the right elevation: The basement is made of stone while the two upper floors are in brick. This type of building is highly resistant to the multi-family condition because of its vertical distribution.
22
after amnesia
Fig. 4 Sana’a: Tower-houses aligned along an urban garden.The effect is quite picturesque, but the house-tower type has trouble aggregating into a continuum.
and significant events are not the same thing. While, a “Do Not Enter” sign conveys a clear message, the gates of a city convey only an ambiguous, stratified, and textured one.16 In pre-modern or pre-industrial societies, traditions and customs ultimately lent a conventional, symbolic, or ritualistic value to what were purely constructive structures. This explains how the emphasis on symbolic meaning led to the over-decoration of door knockers, door jambs, and arches over the doors in houses of the Maghreb. The code, to borrow from the linguists, was overburdened, while the structure remained unaltered and, to force the analogy further, the code and the structure could exist separately.17
the fourth typology
23
type and the modern standard A comprehensive definition of type must also distinguish it from the modern concept of standard, which refers to a precise code. A standard, for example, establishes the height of buildings proportional to the width of the street without involving the system’s other components. Its abstract language (formulas, indices for the suitability for buildings to the public health, proportional ratios) is difficult for the ordinary person to understand and lends itself to arbitrary interpretations. On the other hand, its apparent simplicity appeals to the bureaucracy that employs it as an instrument of power. Moreover, it is indifferent to place and therefore easy to export, unlike a type, which is anchored in the experience of the constructed object and modified in time and place. Argan incisively captured this difference when he wrote: “A standard is not a kind of form but a kind of object: a tool, a machine, equipment, a house, and if you will, a city. And as such it takes the place of the module in the process of classic planning to the extent that one can claim that the greatest discovery of modern architecture is the substitution of the module/object for the module/measure.”18 The confusion between type and standard is part and parcel of the history of the Modern Movement and explains the aversion that so many of its greatest exponents, such as Aldo van Eyck and Giancarlo De Carlo, had to typology. When De Carlo accuses typology of causing social segregation, and types of being “so rigid that the invention of alternatives seems useless,”19 he is really talking about standards. In a way a standard is society’s mistaken reaction to the progressive loss of its own collective memory of construction. The many standards and indices attempt to compensate for the loss of quality of space by exerting numerical and quantitative control. During periods of the kind of conscious spontaneity that characterized all of the Mediterranean region in the Middle Ages, there was little need for regimented norms and standards. This is obviously hypothetical, for even in the Arab city, where good everyday relations between neighbours of the same religion is assumed, there was nevertheless a certain regulation of public spaces (what they were varied according to the regulation of public spaces as dictated by the four legal schools). The same was even truer for Italian medieval cities where periodic attempts were made to reopen public streets by ordering the demolition of staircases, balconies, and overhangs. At the same time, the idea of the common measure was instituted - both Rome and Florence had its own distinct set of measurements. These systems of measurement were among the highest expressions of Italian civilization in the Middle Ages, publicly proclaimed by city governments to their citizenry, and they remained in effect until the introduction of the metric system in modern times. Thus the periods of conscious spontaneity in building were not without their measures or standards, but they were far more easily embraced by the citizens, because they were incorporated into the idea of type. If rigid standards were not needed in periods of conscious spontaneity and cohesive society, logically one might expect the opposite to be true for
Fig. 5 The Vitruvian triad of firmitas (technology), utilitas (layout) and venustas (form) is integrated into another triad: architect, client, craftsman. The two triads have a point in common in firmitas-craftsman.
24
after amnesia
periods of crisis or stagnation, that is, standards would be used exclusively. However, this is not in fact the case, because remnants of the idea of type continue to linger in the collective memory of both the designer and the ordinary citizen. A notable example is the kitchen presented at the Frankfurt CIAM in 1929. It so influenced architects that it is considered a design landmark, which changed kitchen design the world over: “...a kitchen was presented which was so well organized and well equipped that a woman could have made an omelette using the greatest economy of motion … many other architects followed the same set of criteria in designing other parts of a dwelling: the bathrooms, bedrooms, living and dining rooms. Then using the same standards, they designed entire homes, buildings, and finally, entire neighbourhoods...”20 But Giancarlo de Carlo shrewdly pointed out how an incorrect understanding of the concept of a type had led to an inversion of subject (the housewife) and object (the omelette). Unlike the architects, ordinary people were much less influenced by this project. One imagines, for example, women who work in the home who recall a type that leads them to point out the non-machine-like functions of the kitchen - its key role in family life, for example, and its function as the heart of the house. For them, the kitchen is not only the application of a set of standards. It should be flexible in size and functional, but it should also be a place where many other activities can take place, such as children doing their homework or domestic chatter. The act of cooking is not simply functional; it is a domestic ritual. In other words, allotting a space and calling it “kitchen” is the result of the specialization of the rooms in a house, but a real kitchen retains intact the role and memory of the remote mono-spatial house. One thinks with irony of Peter Eisenman, who in 1988 could brag just the opposite, of having changed the eating habits of his client, sprinkling the dining room with columns.
the fourth typology
25
the a-priori type Neither is it entirely impossible that a shapeless piece of marble or stone left to the incessant movement of the sea should one day look like something else, the face of Apollo. That is, if a fisherman is familiar with that divine face, he may recognize it in that marble in the water. Paul Valery, Eupalinos ou l’architecte It is not nostalgia for a distant past that leads us to refer to vernacular examples, but the expression of a conscious spontaneity that they embody. The striking unity of a Kabyl settlement on the mountain above Tizi Ouzou in Algeria (fig. 82) or a ksar in the Draa valley in Morocco, where the houses and overall fabric are notable for their homogeneity, is possible because at a given point in the past every villager building a house referred to the same common legacy of constructive, distributive, and decorative techniques. The owner very likely did not even need to describe a house to a local mason who shared his cultural milieu; he simply told the mason what he needed and the mason built the house without too much planning, and the carpenter, when he built the roof, built the only roof he knew.21 Changes over time occurred so slowly that almost any abrupt change in the formal continuity of any one built object was essentially cancelled out. Homogeneity, far from being monotonous, was the aesthetic merit of the village. This is not to say that differences did not exist. “A shepherd of a large flock who must distinguish one sheep from another necessarily takes into account the subtle differences between his individual sheep,” writes Tessenow, “and he is quite able to do it, while a non-shepherd like ourselves thinks they all look alike because in this case our eyes are not used to seeking out the subtleties.”22 Paradoxically, when the protagonists of the modern movement tried to cut the bonds of historical consciousness and by extension the historic fabric by using a clean, abstract language, they constructed that language from the southern Mediterranean vernacular. Masters of modernism such as Joseph Hoffman, Adolf Loos, Adalberto Libera, and Carlo Enrico Rava, for example, were inspired by examples of Libyan, Tunisian, and Algerian vernacular building.23 A patrimony of expression of space and architecture still exists that people retain in their memory and apply when they build a house without an architect. If we visit a gurbi an illegal settlement on the periphery of Tunis, such as the Melassine quarter, or a douar at Marrakesh, places where we would expect chaos, we will be surprised by the rationality of the layout, with its equal-sized plots properly aligned along regular streets. Most striking is the similarity of these layouts to the fabric of the madina. The immigrants from the hinterland, who have occupied the land and subsequently built these structures, have used the patrimony expressed by the type simply because it is ingrained in their consciousness. These
Fig. 6 View of the historical centre of Constantine, Algeria. The architecture of the city appears as the continuation of the rocky facade of the canyon. Today the continuity of the fabric is punctuated by an ever-increasing number of voids created by demolition.
26
after amnesia
and other similar examples reinforce the principle of type that Saverio Muratori called “a common creative effort.”24 His observations on the built fabric of Venice and Rome also led him to the important and original corollary that type is not only an a-posteriori mental construct, but it already exists within the built reality of the building, the fabric, and the city. It is “the mental project” of whoever builds or remodels a building, and therefore precedes the planning stage as a pre-representation. Furthermore, Muratori points out, at their deepest level, types are much more than just schemes established a-posteriori. They are the essential formative elements of stylistic forms and also of the works of art themselves. In these works they represent the often decisive contribution of an environment and a culture operating at the individual level while characterizing an entire school, age, and people. Outside such a formative spiritual climate single works would not even be conceivable.25 Unlike Platonic ideas, the a-priori type does not pre-exist at the metaphysical level or on a formal schematic level as Argan suggested, but is a product of the historical process and is rooted in a society’s culture, as John Habraken never tires of stating.26 The a-priori type is a sort of mental project which is stronger than Donald Schon’s27 notion of reference. For Muratori, the type is the concept, not a scientific paradigm, a conjecture with which to verify the sensible world except in contrary cases; rather, it is scientific in that it exists and has its roots in History. If it were not immanent in reality/History, which helps to interpret, it would in fact be a broken tool and we would make the usual error of constructing an abstract and subjective system like the gestalt of Christian Norberg-Schulz, entrusting ourselves to theories based on perception. It is neither the work of individuals nor of a society in a given historic moment, but is slowly formulated and progressively added to by society as a whole during its cultural-historical evolution. Since it is formed on the structure of the environment and on principles and structures of use as experienced, the apriori type is deeply tied to the place and is opposed to the conventionalism of standards, but to the atopic as well. It is always politically, culturally, and economically up to date. While it is shared it is also individual, insofar as each person who uses the type introduces new elements that make changes in it that are not part of the existing consensus. We may sum up by saying that an a-priori type is determined by the legacy of transmittable characteristics which precedes the formation of the single building, governing its structure of relations from within. In other words, it is the body of customs and norms acquired over the course of the building experience, which forms the framework for previewing the proposed building.
the fourth typology the a-posteriori type If the object is a spider web then one knows a spider made it; a honeycomb then bees built it; a mole hole then a mole made it; and a nest is built by birds. John Ruskin In today’s world of global communications, spontaneous societies that have preserved a coherent homogeneity and can generate a type a-priori have been marginalized and function mainly as subjects of anthropological studies.The a-priori type no longer exists globally, but only in kaleidoscopic fragments. It can only be described a-posteriori, that is, through a critical synthesis of the historical record, including the fossilized fragments of the people’s unconscious.This perspective complements and is consistent with the previous definition of the a-priori type. At the same time it does not contradict the definition of the a-posteriori type as “an organic sum of the invariant morphological features of a group of buildings from the same time period and cultural area,”28 where the term “organic sum” refers to the patient untangling and critical reassembly of scattered and disorderly fragments. The similarity that exists between the taxonomic procedures of some scientific disciplines and typology’s a-posteriori procedures for disassembling is purely illusory. Both the strategies and the objectives of these other disciplines are different. Taxonomic procedures aim at finding the differences between objects in order to group like objects into classes and species, as in the work of the entomologist or Calvino’s sand collector.29 Typology, on the other hand, must take a historical approach to the similarities or structural bonds between objects to locate their common roots. According to Muratori, “The type as such is no longer the standard sequence of positivism nor the fixed biological variety of determinism.” Even in 1946, when he wrote these words, Muratori recognized the need to insert type into the complex flow of history, intuitively connecting type with the concept of organism and embedding both in society. In fact, he continues, “the architectural type is a kind of architecture and therefore a building organism which, as a result of repeatedly taking shape in order to respond to the typical needs of a given society, ends up so intimately adhering to its psychological climate that it absorbs all its essential human traits.”30 Type, then, is not merely a sub-product of the historical process that leads to the mechanical repetition of needs or material development. It is history. The connection of type to historical process is the only correct use of it in the design process, since the type is an authentic expression of the collective memory. One of the many exceptions is neoclassical architecture, where the opposite is true: although based on historical forms, it is ahistorical, because its designs stem from the formal classification of types that assume classical schemes, and not classical architecture, as models. Due to a rupture in the historical process, at present type can be approached only through a reflective critical consciousness. This operation of critical consciousness to reach spontaneous consciousness that will, at best, fall short of the goal, is what Muratori called “asymptotic.”31
27
28
after amnesia
interpreting There are some things which ordinary people never see, not because they are secrets but because they have never been pointed out to them. Alessandro Giannini, La periferia e il progetto The universal diffusion of writing, which appeals more to the imagination than to reality, has in part cancelled out the capacity to read other evidence, such as architecture and music, which have produced nonverbal objects just as full of thought. Deciphering the iconography of a building, of an aggregation or a territory with terminology borrowed from philology, that refers to written texts,32 here alludes to form, the tangible and living expression of the structure of things and their order, form as legacy or mark (whence formwork). In spite of the abrasions, signs and additions, as in a medieval palimpsest, man’s manipulations on the earth’s crust always in fact leave a trace that cannot be confused with the actions of nature. Reading signifies interrogating the geometric representations of a building, the iconography of a city or the cartography of a territory in their discrete elements, recognizable either by their recurrence or the uniqueness due to the emergence of the form. Reading involves a critical analysis that proposes to identify a-posteriori elements and structures as well as reciprocal relationships; it assumes that all man-made reality possesses internal laws and relationships. Disorder cannot be read, not because it lacks an internal order, but because our interpretative tools are not adequate for reading something so complicated. Structures and relationships are not always visible (for example, in Roman and Byzantine wall systems the structural nodes have been plastered over), but they can be logically deduced both from what can be seen and by analogy from other examples that identify the same type. Reading can at times be fragmentary, since fragments give back the text, but it can not lose sight of the overall unity. In reading, it therefore makes no difference which scale is focused on first, since one will illuminate those next to it, and by the same token, the reading can be interpreted to introduce parameters and data that can help clarify this. If the scale of application is urban, for example, at a certain point the - introduction of territorial parameters such as the interference of large routes, their entrance into the city and the permanence of signs in the urban fabric can be essential for opening up the analysis of the foundation of the city. The purpose of reconstructing the process by which types are created is to regain the ability to understand these structural relationships that were negated by the vanguard of the Modern Movement. Regarding the propositions that guide us, it is useful to first state some important concepts. The depth of analysis of an a-posteriori type ranges between a maximum and a minimum of typological specificity: the level required in distinguishing a house from a tree, for example, is minimal, but the level needed to distinguish a row house from an apartment house
Fig. 7 Resolution in graphic representation is the quantity of information that a certain scale can indicate. A shows the streets and culs-de-sac; B the land subdivision; C identifies the first qualitative attributes of the buildings on the plots by showing the building footprints and courtyards; D deta ls the interior organization and surveys of ground floors.
the fourth typology is much greater. The ultimate theoretical level of typological specificity is the category that consists of only a single building, and building type and building become one and the same. In this particular circumstance - which exists only in the mind of the people and under spontaneous conditions the a-posteriori and a-priori types obviously converge. Knowledge about the level of specificity of the type allows us to evaluate the quantity and quality of characteristics to introduce to the reading: for example, in order to apprehend the urban fabric of Cairo one need not study the roofs of every building.33 The concept of a hierarchy of scale complements the concept of a select specificity. As Durand proposed long ago, in the relation of the parts to the whole a sequence of relationships is established in the built object. The architectural materials are first organized to form structures such as walls, framing, and roofing. These in turn are used to form rooms, the rooms grouped together to form buildings, and the buildings grouped to create blocks, and so on until a city is built. Unlike Durand’s combining mechanism we must take into account that for every rung on the hierarchical ladder there is a corresponding increase in the level of complexity, if only because of the greater number of elements involved. To say that one builds a city out of columns is an aphorism comparable to that of Leon Battista Alberti’s comparison of the city to a big house.34 An important corollary emerges with regard to the representation of parts and structure in the principle of level of resolution, discussed by Karl Kropf35 and defined as the number of parts that can be included in the representation of a structure. In the plan of a city block, for instance, one begins by representing only the outer streets. When one adds internal property lines a higher level of resolution is achieved. An even higher level is obtained by adding building types and a higher level than that by adding specific details such as the placement of stairwells and doors. By this time we have matched the superb quality of the French cadastral maps for Aleppo scaled 1:500 (fig. 110). One always needs a verification through direct fieldwork; many phenomena can in fact be read directly on the facades of houses, from the placement of openings, from the renewals of the wall, etc. In the l960s Muratori defined a “universal categorizer” for obtaining a detailed and unified picture of reality. This universal categorizer is based on a progressively more complex system of logical categories, organized on a Cartesian grid. This philosophical endeavour sought to evaluate the cultural productivity of man by comparing the capacity and resources of the object and the needs and attitudes of the subject. This appears as a dialectic of questions and answers where questions correspond to subjects (or goals) and answers to objects (or means). In architecture, according to Muratori, the subject process passes through phases of progressively deeper evaluations of reality while the object process passes through progressive phases of resource utilization. In Muratori’s view the progressive phases are a function of the way in which man sees the world and, more precisely, how he judges it logically; how he uses it economically; how he evaluates its social attitude ethically; and how he evaluates its expression aesthetically.36 It is an important theoretical legacy that opens the way
29
30
after amnesia
Fig. 8 Chart for the classification of buildings at the first level of typological specificity. On the horizontal axis the various levels of scalar complexity are plotted; on the vertical axis the structural complexity is indicated. (From: Paolo Maretto, Nell’Architettura [Florence, 1973]).
to numerous applications in the analytical realm: the general table in the version edited by Paolo Maretto appears with some terminological variants in figure 8. Muratori’s philosophical endeavor draws from the Hegelian subject-object dialectic and Benedetto Croce’s re-interpretation of idealism.37 Muratori’s thought remained entrenched in a subject-object duality. Today such a traditional contrast between subject and object, man and reality, can be enriched either from a phenomenological perspective or by articulating it in terms of Popper’s epistemological “theory of three worlds.”38 According to Popper, there is a body of theory that mediates between the physical world (object) and the world of internal assimilation (subject). Popper thus provides a theoretical role for the typological discipline that stemmed from Muratori’s philosophy. For that reason Muratori’s prophetic visions should be evaluated from a historical perspective. His philosophical framework in the context of architectural studies leads to the table in fig. 8. The hierarchy of the subject’s complexity is on the X axis, while the scale hierarchy of the object is on the Y axis. The complexity of the subject is subdivided into the element, the structure (of elements), the system (of structures), and the organism (of systems). Elements are the simplest components of the organism and correspond to the building’s construction material: the brick, the beam, the column, etc. Structures are composed of elements bound by a relation of dependency and recognizable by a coherent geometric form, for example a frame. Systems are in turn composed of structures connected by a relationship of necessity and dependency. Structures can be isolated by their specific function in the organism. Organisms are structures of systems with an autonomous character, for example, a building. In a traditional organism, three principal systems can be separated in the interpreting: namely, construction or mechanical systems; layout or distributive systems; and morphological systems. In the same fashion, the scale hierarchy of the
the fourth typology
31
object is subdivided into house, neighbourhood or aggregation, city, and territory, categories which will guide us through the scalar passages of the reading in the chapters that follow. We can as a first reading apply Muratori’s guidelines for a critical a-posteriori reconstruction of the overall building reality and impose a proposal of synchronic reading with a double entrance matrix, proceeding in the following stages: 1.identify the scalar structure, i.e., house, neighbourhood, city, territory; 2.identify the structural complexity, i.e., the element, structure, system, and organism; and finally 3.find the shifts in scale -either up or down- and the way organisms can become basic elements in the formation of a new type on a new scale.39 The scheme of Muratori’s table is deceptive. In spite of its appearance, Muratori conceived each quadrant as open and divisible into multiples of sixteen for the purpose of pursuing greater and greater detail, in order to more closely approach “awareness.” We shall use the table conventionally, keeping in mind that the grid scheme is a guide for thinking. The shift from the general table to a first level of specificity is represented by yet another further division into sixteen modules, and hence four levels of scalar complexity multiplied by four levels of structural complexity. Each sub-multiple (divisible again into sixteen submultiples) corresponds on the abscissa and the ordinates to the characteristics of typological specificity for each scale. For example, the abscissa of figure 9 corresponds to categories
Fig. 9 Typological specificity on a higher level of complexity. A single quadrant of the first chart subdivided into sixteen frames. The arrow represents the leap from the urban individual to the type, or when the characteristics of the specific building, specific block, specific fabric and specific territory coagulate into the building type, into the typical fabric, into the typical settlement and territorial type.
32
after amnesia
of urban fabric specified as linear or crossed, continuous or discontinuous, while the ordinates represent the categories of urban units specified in mono-directional or poly-directional connections. The operation of opening the quadrants and the consequent increase in typological specificity reach extreme, even audacious levels of approximation. In Civiltà e territorio, Muratori demonstrates this by developing his Table 1 into 156 squares.40 Others have attempted similar operations in different scales. Giancarlo Cataldi’s innovation was to introduce time as a parameter onto the abscissa to create a sense of process which was relevant to his research on the primitive house.41 Figure 9 shows the reasoning about building structures in the level of structural complication between system and organism. The nodal moment of the synchronic reading is the shift from the last (bottom) square of a column to the first (top) square of the next column. This represents the shift from unique individual organism (defined by the harmony of its components) to a typological moment of the next step when it is possible to isolate the common characteristics of buildings belonging to the same time and place from a mass of individual buildings and define them as a type. This coincides with the route of the method that we will follow, dedicating a chapter to each level of ascending complexity. Put more simply: the cognitive process of an ascending level proceeds through successive passages of structural complication from a moment in which the materials - or better, the material in which we recognize a predisposition to building - are joined in structure and then in system until the building organism is realized in a specific building individual, in which a specific family is identified. The logical passage in the successive cycle, which corresponds to the level of the unit, occurs when the individual building loses its individuality to become valued for its way of being in function of those similar to it, and it becomes type. The building type at the base of the aggregative law implicit in its nature joins with other similar types running through the entire graduated cycle of the fabric, that is, building fabric and street layout, finally ending up as the aggregate individual, the quarter or neighbourhood in which a specific extended family group becomes aware of itself. The logical passage to the next cycle presupposes the recognition of the typical characteristics present in the neighbourhood that becomes a typical urban nucleus, the module which, joining together, blankets the whole city. The urban fabric grows until the free spaces have reached the saturation point, generating between each module special types of linkages we call nodes. These denote the variation from a purely quantitative to a qualitative growth. The urban organism qualifies as a city, or urban individual, when it reaches a complete formal definition and its nodes are transformed into monumental emergences. At this point the phenomenon has become self-aware. In the territorial cycle the urban individuals become settlement types or minimal elements of the territorial scale. These are places in a dialectical relationship with the land fabric that constitutes the modular structure of the territory, while the road system of the connections takes on the unifying qualitative function. A territorial organism reaches conscious self-determination of the territorial individual
33
the fourth typology
Fig. 10 Example of frame 5 in fig. 9, comprising the different structures and their respective materials (From R. Bollati, S. Bollati, A. Giannini. Quadro generale delle strutture architettoniche, Rome, 1978).
when it succeeds in giving itself precise borders and significant emergences in an ordered hierarchy. This process entails not only shifts in scale but also in concept. Some important corollaries follow. First, the domestic building type becomes a formative element of structures and systems on a much larger building scale. Second, the ability to aggregate is confirmed as a conditio sine qua non for the urban nature and sociability of a type. At the end of the four cycles on the scale hierarchy described above, one would have isolated a territorial organism with all its components and multiple relationships, but the process is synchronic and our task only half finished. The task of typological analysis is not to verify a-posteriori several common characteristics for classification - although such a task is useful for understanding the structural relationship of parts - but to reconstruct a sense of the changes in a type through history, that is, through the built environment. As Aristotle noted, the essence of a thing is found in its changes: the essence is interpreted as the sum of a thing’s inherent capacity for change, and the changes are interpreted as the realization of this capacity.
34
after amnesia
serial and organic structures In the relationship of parts to a whole, the components of a structure have differing degrees of interdependence according to their complexity and their ability to be broken down into other elements. Depending on the structure of the elements the number of possible groups ranges between the two extremes of serial or organic structures. Take the case of a campground: the relative position of the tents can be changed and even eliminated without altering the nature or form of the system. In the case of the human body the opposite is true: each element is distinct from the others and highly specialized. The only way for a head or arm or leg to have a life of its own is as a part of the hierarchized whole. It follows that not only are the parts not interchangeable but the very life of the organism depends on the existence and position of the parts. The single parts of the whole have a relation based on necessity and mutual dependency. In the case of a tree, if some of its roots are cut away it will compensate for the loss by reducing its foliage. Before undertaking a reading, the concepts of organic and serial need to be considered because they relate to the question of local conditions and explain how a single component - an arch, pillar, or even a house - can be combined on different levels of seriality or organicity in different architectural cultures and is not simply derived from climatic, economic, and technical factors. Saverio Muratori astutely associates a more or less organic method of constructing architecture with a more or less organic way of constructing thought.42 Using such a conceptual framework, similar structures in different areas can appear very different. For example, Amos Rappaport claims that as an archetypal principle of enclosure the courtyard house is essentially the same all over the world. But this similarity proves fallacious under closer scrutiny, as one sees when looking at the various ways in which environments within the enclosure are grouped and distributed in China, for example, as opposed to the Maghreb or Pompeii. Therefore “organicity” or “seriality” can almost be regarded as a cultural trademark.43 In China’s courtyard, the individual elements tend to remain separate along the sides of the enclosure; even in cities under strong demographic pressure the corner rooms are not integrated to form a unitary whole. The exact opposite occurs in the peristyle of an ancient Roman domus, where what could have been a series of rooms tended instead to form a unit. The Maghrebian patio house is a sort of happy medium, where any irregularities of the lot are absorbed by the perimeter rooms while the regularity of the empty center space is organic, albeit reduced in size by the position of rooms that interlock like dominoes. The temple dedicated to Confucius at Qufu is probably the most important temple in China. It consists of six pavilions lined up along a north-south axis. The entrance is marked by a monumental arch and the last pavilion, Dacheng Hall (the repository of a statue of the philosopher), is twice as tall as the others. A minimal hierarchy indeed exists, but the general impression is one of grandeur achieved through the arithmetical summation of many equivalent elements on a ritual axis. Ritual alone
the fourth typology dictates the position of the elements that are otherwise as interchangeable as the numbers in a sum. Thousands of kilometers to the west of China, Mimar Sinan demonstrated the organic quality of Ottoman thought. He borrowed the layout of the Haghia Sophia for his Shehzade Mosque in Istanbul: a square enclosed by two semicircles and capped by a central dome and two semielliptical domes; each room of the two access porticoes in turn are capped by smaller domes. The whole roofing system is regulated by a global concept of space in which the nodes - the connections, squinches, and pendentives - in every corner suggest continuity. The visible supporting structure of large columns inside and powerful towers and angular pillars outside emphasizes the continuous distribution of forces and the interplay between parts of the organism, legible in the section. The nature of a material also plays a role. The pavilions of the Qufu temple are masterworks of carpentry; the Shehzade is a masterwork of masonry. Materials employed by a culture for building purposes have typical characteristics which influence the degree to which a structure is serial or organic. Elastic materials (wood, iron, and steel) are linear in form; one dimension dominates the other two. The materials lend themselves to repetition in a series. Structures made from them are therefore strongly serial, that is, they lend themselves to groups of discontinuous structures through repetition. Plastic masonry materials (stone and brick) are in contrast characterized by two dimensions dominating the third, that is, the surface. The materials can be made continuous and structures made from these elements are strongly organic and mainly homogeneous. For example, a wall of bricks is as homogeneous as a wall of concrete or mud. The position of a structure within an organism determines both its geometry and size to the extent that if any element is substituted, the organism loses its identity and function. A further distinction needs to be made regarding techniques of construction: when using timber, the supporting skeleton of the building is constructed before the closing structure which allows the serial process to operate. With masonry, on the other hand, the building is closed and covered at the same time and is therefore more naturally closed by a vault or dome rather than a flat roof. In the Mediterranean, ancient Rome introduced a masonry culture which employed the arch and the vault. The ancient Greeks, on the other hand, preferred post-and-beam frames. The Greek temple thus is the product of a timber culture using plastic materials, although the material is stripped of its basic grammar and construction techniques. When a serial building tradition begins to use organic material, it tends to alter the nature of the material. For example, in the ancient Roman building in Djemila (Algeria), the vertical masonry wall is framed by stone posts at regular intervals, a serial use of an organic material (probably of Phoenician origin) that represents an alteration made by a building tradition that once clearly employed timber and canonized the rules of timber construction. At times, the two building civilizations coexist, as at Bukhara, where all the monumental works built of bearing brick use a vocabulary from the
35
Fig. 11 The roof of the Karawiyin Mosque in Fez, Morocco. The naves parallel to the qibla reveal a strong seriality based on the rhythmic repetition of the Arab mosque, but the odd number of bays and the larger dome on the entrance bay create a central axis and establish a hierarchical order suggesting an organic layout (see also plan in figure 159.1).
36
after amnesia
plastic-mural world, while the courtyard houses, although still covered by dirt plaster, reveal both in their bearing walls and in their orientations a vocabulary taken from the elastic-wood world, which seems to have been imported from central Europe. An even more obvious example is the redsandstone city of Fatehpur Sikri in India of the 16th century. The building tradition of timber post-and-beam construction, which uses the skills of a carpenter, was adopted to stone and masonry material without modifying the principles of timber construction.44 Two of the above examples also have exceptions to their generally organic or serial structures. The Dacheng pavilion at Qufu is differentiated from the others by a vivid red roof and an enclosing fence. In the Shehzade, rather than rest the elliptical half-domes on a round base as in the Haghia Sophia, Sinan lays them on three rectilinear sides, rupturing the fluidity of the Byzantine space. These examples demonstrate that the built environment is never completely serial or completely organic because objects are born of the encounter between serial situations and organic interventions and vice versa. Fatehpur Sikri and Djemila demonstrate that the natural propensities of materials and elements which define the character of an object can also be altered by cultural uses. This suggests that four rather than two fundamental systems of aggregation exist: two are the result of a clear and consistent relationship between the parts and the whole and the other two of inconsistent use, viz. systematically serial, that is, serial elements grouped in a serial whole; totally organic, that is, organic elements grouped in an organic whole; occasionally serial, that is, serial elements grouped in an organic whole; and episodically organic, that is, organic elements grouped in a serial whole.
Fig. 12 Shehzade mosque in Istanbul by Mimar Sinan and Hagia Sophia compared. Note the different transitions in the lateral semielliptical dome.
Fig. 13 The wagon ceiling of the central hall in the houses of Abu’l Fadl and Faidi, Fatehpur Sikri, India, are typical examples of an organic space realized in a serial area of predominantly wooden construction. Note how all stone elements are prefabricated and embedded (following a technique called ladao) by employing woodworking techniques.
the fourth typology
37
Fig. 14 Example of frame 9 in figure 9 comprising the different layouts and their respective structures From R. Bollati, S. Bollati, A. Giannini. Quadro generale delle strutture architettoniche, Rome, 1978).
The maximum result of an aggregate is realized in the first two cases, where agreement between intent and capacity is greatest. The principle of maximum efficiency in the serial or organic categories is derived both from economic considerations and from historical reality. This is best illustrated by Caniggia’s paradox: I am quite aware that the capitals of an ancient temple are not appropriate for building a wall; they are shaped in such a way as to perform a particular function and assume a specific position in order to resolve a particular building dilemma: how to connect an abutment, the column, with a horizontal element, the architrave. It is therefore conceptually wrong to build capitals on a wall. It would involve using highly organic elements in a serial system which would be better served in terms of both cost and effort if made, say, with square hewn stones. If, however, I lived in the fifth century and my basic problem was to erect the walls of my city quickly with the materials at hand in order to defend myself from an invasion of barbarians, and if I had temples at my disposal - less useful to me under the circumstances of needing to defend my life and property - then I would do well to dismantle them to build my walls.45 On the other hand, this is how much can be touched by hand in the temple of Ba’al in Palmyra, transformed into a fortress in the Mameluk era: the walls were built with the shafts of the temple columns, placed horizontally, so that the entire wall appears to be a heap of trunks; in the wall of the citadel of Ankara, on the other hand, classical sculptures act like blocks of bearing stone.
38
after amnesia
the typological process In its countless alveoli space contains compressed time.That is what space is for.The entire skeleton of the city becomes its collective memory through the spatial memories of its constituent parts. Malcolm Quantrill, The Environmental Memory Introducing the notion of time into the analysis of urban phenomena not only results in giving life to the type we have defined up to this point, but also serves to fix its mutations in history. What is time vis-à-vis urban history and typological analysis? Historians take positions between two opposite poles. Sociologist Georges Gurvitch underscores the distinct temporalities which belong to different histories. Gurvitch’s temporalities are many: “time of the longue durée and slow motion, time the deceiver and time the surpriser, time with an irregular beat, cyclic time running in place, time running slow, time alternating between running slow and fast, time running fast, explosive time.”46 Other historians privilege the precise moment of every individual masterpiece, and so package histories of architecture that are a collage of single events like “the moment of silence in the ticking of a watch,” in the words of Kubler.47 Time for the typologist, as well as for the French Annales school, is slow and drawn out: as a result unusual examples will be referred to as part and parcel of a sequence of underlying ordinary causes, the long drawn-out time encompassing much more within a period. Our idea of “type” as history cannot be separated from the idea of process. The most relevant concept of typological theory is that of
Fig. 15 Schematic representation of the typological process of residential buildings at three scales - building type, building tissue, urban fabric. The diachronic process is the mutation of the leading type and occurs in a sequence of phases. In the figure below, for every leading type a limited and indicative number of synchronic variants are shown according to their position in the block, topography, etc. All synchronic variants slowly lead to the formation of a new leading type that can be
identified only after remarkable transformations have become apparent. The leading type in a given period can only be found in the contemporary developments of that period, for existing and consolidated urban fabrics can only accommodate small-scale mutations. In the figure, crises are indicated by vertical cuts. The specific case corresponds to the period of the Black Death in Europe (1348) after which the city no longer develops unt l the nineteenth century, but changes its fabric through adaptive variants. At the end of the crisis, when the city begins to expand again a loss of spontaneous conscience begins to surface, and the leading type is recreated by merging local synchronic processes with bu lding models imported from leading cultural areas. The typological process continues at its own pace despite the break with history claimed by the Modern Movement, at least in Western Europe, although at present the urban fabric seems unable to recover from urban ruptures.
the fourth typology processuality, which fixes the mutations of the type in the historical duration. Without it, the type runs the constant risk of falling into one of the historic errors: an ahistoric formal scheme good for every trick, or a mechanical montage of forms broken off from the real world. It is necessary to understand the internal mechanism that animates the type and anchor this in built reality, which would otherwise not be different from any biological organism. This is the most relevant concept of the method of typological analysis. Without it, the type continuously risks being misunderstood as a formal ahistoric schema or as a mechanical assemblage of forms. By the concept of process we understand the internal mechanism that animates the type and anchors this mechanism in the built reality. Typological process allows us to understand the evolution from one type to the next, but does not really deal with the question of dating (the province of traditional historians) so much as it does the question of the sequence of buildings and urban fabrics. In fact, similar typical conditions can be isolated in the history of different civilizations in different periods. One example is the transformation of the courtyard house into the multifamily apartment building. In Italy, it was a phenomenon that began at the time of the Roman Empire; in the North African medinas it took off only recently after the independence of the respective countries. The tracing of typological processes allows us, on the one hand, to establish those characteristics of the building that are essential for the determination of continuity in the process over the course of its transformation. On the other hand, it also determines those characteristics that constitute departures or exceptions, and in their own way contribute to the valuable heritage of experimentation. In other words, typological processes show us at the same time both the rule and the exception. “Phase” is defined as the period of time needed to allow the clear identification of changes in the built object. The progression of phases makes up the diachronic typological process from the Greek dia=through, and kronos=time,48 if conducted in a culturally homogeneous area that is, with negligible external influences, such as a closed valley, are syntopic from the Greek syn=together, and topos=place. Before proceeding further we must distinguish between the basic buildings and the special buildings: by basic buildings we mean all houses that are an expression of the primary need for dwelling, and by special we refer to those like the convent, the mosque, hospital, theater, that stand out from the built context, those elements of the collective function and urban qualification that we call facilities or monuments. The aristocratic palace of the Renaissance will therefore be considered a special type, since administrative and representative functions are more important than the dwelling function, equal to the rab, a communal living structure widespread in Cairo, where it exists in symbiosis with the wakalas, spaces for commerce and small production activities, without otherwise being influenced by them (fig. 72-73). The opposite is true for a basic type of the dynastic courtyard house of the Maghreb city. In Fez, for example, the richest dwellings, realized through combining various lots, lack the symbiosis and change in scale to a monumental courtyard, and maintain the layout in several patios with separate functions. The result is visible in
39
40
after amnesia
the very homogeneous grain of the building fabrics of the city, where the special buildings emerge exclusively (fig. 159.1). In Algiers, on the other hand, the urban buildings represent a jump in scale and the passage to a higher type. We understand the reason for this since the time when, in the Algerian capital until 1830, the dominant Ottoman class is different from the local one. The basic type in any typological process coincides with the ideal house in any given time and progressively changes through specialization from elementary matrices to complex derivations. The type is commonly recognized by every inhabitant, and it can accommodate slight changes based on its role and source of revenue, provided they are within the bounds of the type. We can also call it a “leading or current type,” since it is the type all members of a society recognize as optimal. In a given phase it can be coherently found in the corresponding building. A synchronic variation is a type of house realized under less than optimal conditions. These can be the result of topographical problems, or of problems with placement in a block or placement in an incongruous fabric. Even under the best conditions there is always a chance for synchronic variations to develop within a group. Take the classic example of a line of rowhouses with vaulted horizontal structures: the lateral thrusts of the vault are absorbed by the walls of the neighbouring house. If there is a slight rotation in the tissue, as is common, there will be at least one house with walls that are not parallel. Such a house will face problems such as trying to raise a vault on a trapezoidal plan and the difficulty of furnishing spaces with odd angles and will inevitably lead to variations in the type. The typological process is as complicated as the urban or territorial organism in which it operates; it more or less involves the intersection of different processes. It is therefore necessary to reduce the complexity of a building type or a contemporary urban tissue by assuming that they have necessarily absorbed their predecessors and then backtracking to find the simplest form of the type or fabric. This search does not require a return to mythical origins, to Laugier’s hut. The elementary matrix is the first documentable type at either the substratum level or an archaeologically measurable level upon which the reading can be based. A diachronic and syntopic typological process (see fig. 15), limited within the bounds of a circumscribed culture area and referring to the residential type, can be described. Taking as an example a medieval city in the Levant or the Maghreb, the “leading type” in the initial phase of the typological process is an elementary type or plan - we can call it founding type - whose dimensions are an expression of that specific building culture. The urban fabric is conceived on the basis of this leading type and is concurrent with it. Modified houses exist, however, on irregular lots or on slopes or located at the beginning of a series, or in a corner and so on; the sum of these experiences generates a parallel process by synchronic variation, insofar as each can cause the imitation of its neighbour, offering itself as a possible solution to a problem. In turn, the parallel processes mature with and modify the leading type as people gain in experience. In the second phase, assuming the continuous growth of the city, the next leading type will evolve by exceeding its limits and refining some
the fourth typology of its parts. In new growth zones of the city the new leading type adapts to the tissues planned specifically for it, and is found mainly on principal or matrix routes or planned routes. This is not the case in the old city centre, where the layout is more permanent because of the resistance of the building tissue to change. What contributes most to the conservation of such an area is that real estate is simply heaped atop the resistant existing tissue.49 In this case the inhabitant must compromise between the concept of a leading type, an ideal expression if you will, and the reality that the building tissue is unyielding. The inhabitant’s intervention will determine two new possible types of synchronic variation. The first are the renovations or mutations made to the elements of the interior without disturbing the main structure; the second is demolition and reconstruction. In neither case will the adaptation of the new leading type in the old building tissue reach optimal conditions. Consider the recent changes in Boston’s Back Bay townhouses, for example: these are three-storey, upper-middle-class, single-family houses that have been subdivided into apartments to meet the changed housing needs of smaller nuclear families. Each floor is divided into one or more units but none of them has the coherence and autonomy of a newly planned apartment in neighbouring Brookline. In this mutation of the Back Bay type, the tell-tale signs of unplanned-for change include a limiting front-or-back-only orientation, convoluted access routes, and the lack of cross-ventilation.50 Variations generate processes that in turn contribute to the development of the next leading type. In this third phase the new leading type is comfortably located in the new yet saturated fabric but it will be subject to modification in the two older building tissues. The more time that elapses between phases, the more difficult it becomes for the leading type in the old tissues to adapt. The assumption that the growth of a city is steady is hypothetical. In practice, after a certain number of growth phases a period of relative stagnation, or even of regression, sets in, often resulting in vacancies, abandonment, and the like. This was typical of all Mediterranean cities after the Black Death almost halved their populations after 1348. In Siena, until well into the mid-nineteenth century, for example, large undeveloped areas still lay within the city walls. A more realistic picture is one in which intervals of more or less intense growth alternate with periods of arrested development and regressions. During the periods of accelerated growth and subsequent slowing down, the behaviour of residential tissues and special tissues is different. The residential tissues easily both accept rapid growth and resist regression. The period of regression affects first and foremost the special buildings, as they represent an investment of cultural and economic surplus by the collective. In periods of economic stagnation, limited building activity inhibits the evolution of a leading type. Where building tissue shrinks, the specialization of the residential type is also reduced from its former incarnations and produces only synchronic variations. This is logical, since the reduced and limited requirements of the surviving population will lead to a simpler use of the old buildings. In Rome after the fall of the empire,
41
42
after amnesia
not only did the population fall from one million inhabitants to 17,000, but all the luxuries and amenities - the circus, the amphitheatres, and baths - also disappeared, and their structures reused as housing. A similar phenomenon occurred in Tripoli and Algiers during the late Ottoman period, and in Naples under Spanish domination: because the walled city was restricted to the area claimed by the walls, it grew in height, blocked open spaces, and turned special buildings into collective residences. In periods of crisis the most adaptable structures are the serial ones, that is, iterative ones; in the special buildings as well the serial parts are those that are most easily recyclable. A classic example is the recycling of a highly specialized structure such as a Roman theatre into residential property. In both el-Djem (in Tunisia) and Algiers (in figure 161) the similarities between the amphitheatre or theatre and the formal configuration of repetitive cells such as used in the row-house type were exploited to this end. When a period of stagnation gives way to a new cycle of growth the notion of a leading type is considerably attenuated in the spontaneous consciousness of the residents and they are able to manage only synchronic variations. As a result, diatopic51 (from the Greek dia=through and topos=place ) formed modifications thrive in the weakened body of the city, and a new leading type is often imported from a distant but culturally dominant area. A new leading type, the result of the synthesis of local processes and the imported model, is then used in the future expansion of the city. Despite cultural differences, this phenomenon is discernible in both Europe and the Islamic Mediterranean. After 1850, the model used in Italy originated in Paris or Vienna but was grafted onto strong local traditions. The local traditions explain why the fundamental typological processes in Genoa, Florence and Rome took such different directions despite their use of the same models.52 In Aleppo after 1870 the Venetian type called a portego was merged with the local type of courtyard house with iwan to produce a tripartite house with a wide main corridor. This model was universally adopted when the new quarter of Aziziye was built in the second half of the nineteenth century.53 Algiers presents yet a different case: instead of agreeing to adopt a single Western model, a variety of types were imported and imposed by the colonizing French population. The impact of colonization minimized the effect of local processes and the leading type coincided almost exactly with the imported model. I say “almost” precisely because even though local building customs were followed only in those residential sections of the city which the French shunned, they nevertheless persisted through many small gestures. Only a detailed reconstruction of the typological process would reveal to what extent they survived. It is clear, however, that the post-1830 colonial construction of Algiers preserved traces of local traditions in the compact dimensions of its building blocks, the result of a minutely divided property substratum and of the demolition and reconstruction of small courtyard houses. The imposing imperial facades of residential buildings also mask a lingering
the fourth typology
43
Fig. 16 The typological process of the Selgiuk and early Ottoman mosque with reference to the three main families: 1. derived from the enclosure; 2. derived from the roofing and 3. derived from the madrasa.
memory of the smaller spans of local traditional modules (see scheme fig.75 above left). There are many other examples but for the moment these are sufficient to demonstrate that even the most ruthless and determined colonial approach cannot obliterate the memory of a place entirely, deeply rooted as it is in local typological processes.54 For special building it is also possible to reconstruct the succession of diachronic mutations in a surrounding area. The role of the current type of residential building is covered by the bearing type, the sum of the characteristics of
44
after amnesia
the civic area to which the building belongs. The mosque of Tlemcen from the 15th century, for instance, or the Merinide mosque of Fez (where the name of the dynasty takes on the characteristic datum of a temporal arc) certainly characterizes a rather general type, which admits a series of synchronic variants that in turn generate parallel typological processes. The two different cycles of the basic and special type are not in competition with one another, but are complementary, since one anticipates the other. The first two examples are a product of a building boom, the others of a moment of stasis. A significant difference resides in the limited number of special buildings with respect to those of the processes of the house. In the typological processes of special building we can read the almost complete gamut of the relative synchronic variants, and archives often furnish documentation in this regard; in the processes of mutation of the residence a reading is possible only in a synthetic way, from the moment that these variations, responding to minimal updates of the type, can be appreciated a-posteriori only as an expression of the diachronic process of the current type. To round out the theoretical picture, at least two other processes should be mentioned: those that have to do with special types and diatopic types. Special types are generated either directly or indirectly by the residential typological process. They can occur over time, but because during some increasing specialization of a house a new procedural line branched off as new, non residential uses were introduced. In the third and fourth centuries CE, one such new branch is represented by the basilica with three aisles (3:6:3 meters) directly derived from the Roman domus, and another is represented by the central plan of the mausoleum. If it is acceptable that the Arab mosque derived from Muhammad’s house in Medina then this represents yet another branching out of the residential typological process in the seventh century in the Middle East, and still one of the two types of Mameluk mosques would be a branch of the process of the qa’a in Cairo (fig. 69 and 70). In the reconstruction of the mutations of the special types, special importance is owed to the meetings and hybridizations that occur between diverse territories (diatopic processes). We must keep in mind the fact that mosques and monuments in general are the work of specialized workers, who in the Dar al Islam would move around in search of learned patrons. What is more, these patrons, like the governors, were rarely originally from the place in question. In principle, if building activity can rebound following a crisis and steadily resume its course without suffering additional setbacks, the leading type of the twentieth century will be a diachronic variation of the earlier phase of the nineteenth, and so on.55 This reality is evident in the urban growth of the 1930s in Europe when Rationalist architects, through the use of a technologically influenced abstract and discontinuous language, expressed their alternative ideology in their urban expansions but nevertheless referred to the leading type of the time. This type, in spite of the experiments performed on it, was nothing more than a middle-class apartment building of the XIXth century (maison de rapport). Similarly it is possible to say that built products in different areas are different, and their differences are directly proportional to the social, cultural, and economic distances of a given time (a diatopic and synchronic
Fig. 17 The madrasa of Qijmas al-Ishaqi (1479-81) in Cairo respects the spatial relationships and the geometric composition of the qa’a of the house and consists of a durqa’a covered by a wooden roof with an octagonal lantern flanked by two iwans placed along the longitudinal axis of the direction of Mecca.
the fourth typology moment). From a broader vantage point the comparison between buildings of diverse territories presupposes the concept of a bearing or leading area, the culturally stronger area in a given time, today we would say compelling, that dictates style. There will therefore be bearing and marginal areas, and at this scale the synchronic variants will be represented by the types of marginal areas, without however excluding that each area has in turn its own bearing type and synchronic variants. For example, in the Ommayade epoch the bearing area is Syria and the marginal area is the rest of the Caliphate. The bearing type of the mosque can be assimilated to the prayer and sahn room type, which we see in the great mosque of Damascus, the synchronic variants of the Umayyad mosque of Baalbek or the first version of the Al Aqsa mosque in Jerusalem. To cite a European example, in the Gothic area the bearing area is the Ile de France; Rome is an extremely marginal area of masonry construction, where the principle of Gothic construction never took hold. This last example allows us to digress to make a clarification regarding the concept of architectural language: we can say that a lexicon is language when it is realized in the bearing area and in those close by, that is, when it identifies with the civic values of those places, but becomes style, an adopted product, when it is imported into marginal areas. Diatopia is a common phenomenon in the Mediterranean when Roman Imperial civilization spread the mare nostrum of buildings of distinctive scale and Euclidean geometry regarded as representative of Roman civilization. Baths, aqueducts, the amphitheatre, the circus, the temple and especially the theatre allowed citizens to see themselves in a familiar landscape, almost like the fast food “cottages”of McDonald’s spread out over the world today. This was possible because of the exceptional coincidence of bearing area, civil koinè and geographic basin for at least six centuries. It is also obvious that houses from the same period in neighbouring areas share similarities directly proportional to the amount of contact they have with each other, a sign that a certain osmosis does occur. These affinities are the result of real influences and should be taken into account by process analysis. The hypothesis that a basic type migrates (it would be a diatopic residential type) is a contradiction in terms because in principle this type is anchored to its social base and rooted in its locale. Nevertheless the importation of a type as a cultural model or its migration with an entire population (as in Algeria after 1830) has already been noted. The Tunisian villages of Little Calabria and Little Sicily after 1890 are further examples of this phenomenon. The villages built by poor Italian emigrants were in fact a bit of Italy abroad. A more complex case is the similarity between types that are encountered across vast geographical areas, for example, the extraordinary similarities between the Venetian, the Dalmatian, the Ottoman Turkish, the Aleppine, and the Lebanese house in the eastern Mediterranean. This cannot be merely a coincidence and might be perhaps best explained as the result of a graft made onto the common substratum of the RomanByzantine house. The contribution of a substratum type such as the domus that exists all over the Mediterranean basin, in an archaeologically permanent form under the medieval fabric and apparent in reading
45
46
after amnesia
through the metrological evidence, is possible. It is not nullified by the many changes made to it, and when encapsulated in the constructive memory of the place it becomes a distinctive cultural fact equivalent to an anthropological deep structure. The concept of a substratum or fundamental type is essential to reading a stage because it allows for the identification of planned tissues which are hidden below more recent spontaneous ones. To sum up: Process is always diachronic, based on continuous time whose rhythm changes in relation to the behaviour of the various scales. The inertia of the large territorial scale with its urban framework and infrastructure produces a slower rhythm, while, on the other end of the building scale, changes occur faster and thus have a quicker rhythm. Differences can also be discerned in the position of the objects on the same scale, for instance between central and peripheral buildings, or between special and residential buildings. The typological process takes place in phases in which the leading type proceeds through moments of equilibrium alternating with synchronic variations. The typological process may be syntopic (i.e., occurring in the Fig. 18 (Above) The leading area and local areas during the Ottoman period. (Below) Atopic objects all generated in Rome contributed to the unity of the Mediterranean in the time of the Roman Empire.
the fourth typology same area) or diatopic (i.e., involving variations between areas). Finally it can refer to the basic type (the house) or to special types. The graphic models we construct in our mind or represent on paper are reductive. Not even the metaphor of a tree with its trunk, branches, and foliage can do justice to the complexity of typological processes, unless it is the banyan tree of India, whose enormity and growth pattern would merit definition as a grove rather than a single tree. Its branches extend out horizontally, shoot upward, descend toward the ground, and sneak underground where they take root and shoot out new branches not unlike the original ones. Eventually only the expert eye of the gardener can distinguish the primary trunk. The reconstruction of typological processes is somewhat similar to the script of a film, with one important difference: the filmed work anticipates scenes which might take place in the future, while the reconstruction of typologies projects into the past. It approaches past scenes with the unquestionable advantage of already knowing many of the elements, relationships, and sequences; it can therefore critically reconstruct the missing pieces insofar as they are typical. It requires the awareness that a critical description of the process -as in all disciplines- must be carried backwards throughout from the final frame. The term “critical” is emphasized: even a reading restricted to documentable facts is not “objective,” nor is knowledge the equivalent of a mass of data piled up in a closet but is remote control operated by our minds.56 This idea is also implied in Piaget’s method of structural analysis. We hear only what we want to hear and nothing else. Of all the possible facts that make up a type, which are the most important? And how shall we proceed to navigate among them? To identify a type is to form a hypothesis which can reach the greatest level of plausibility only through continued inquiry.
47
48
after amnesia
interpreting and design The best artist is not the one who breaks the rules but the one who changes customs, in much the same way as the best player is not the one who cheats but the one who invents unusual solutions during the game. Giovanni Pozzi, La parola dipinta In Popper’s philosophical framework of the three worlds, I think of the architect as akin to Charon ferrying spirits to the underworld: they are condemned to ferry practical experience drawn from the physical world over to the conceptual world through the intermediate world of the imagination and back again against the current of the conceptual to the physical one. In the past, when spontaneous consciousness was rooted in man, the limited and precise role of the intermediate world of invention (understood in its Latin meaning of invenire = to find) was to adapt to customs, norms, and traditions derived from a natural perception of the physical world and deposited in the conceptual world. The role of the world of imagination grows disproportionately in periods of critical consciousness such as the one in which we now live, a period which can be traced back to the Renaissance. Design is inevitable for many reasons: the demise of spontaneous consciousness erases traditional construction skills; the complexity of modern buildings exceeds the traditional masterbuilder’s capacity to control construction, and its realization requires the coordinated efforts of many specialists; the necessary financial investment imposes deadlines that traditional construction sites are unable to respect, and so on.57 There is no returning to a golden age, so it is important to be aware of certain dangers inherent in the necessary activity of design. Design offers both the great advantage of planning a construction in advance and the disadvantage of almost always separating the builder from the inventor. It is an expression of a critical conscience detached from reality and better able to relate to other designs (in magazines perhaps) than to the constructed object itself. Design may not derive from the specific civil context of a certain area, but from the special personal universe of the architect, and its inventiveness may be based on a memory that is without context and exportable. This personal language of design is responsible for the destructuring of local dialects because it favours architectural contraband. Perhaps this is not a new phenomenon for special building types, but its extension to more mundane types and to the residential fabric has been disastrous. Today every individual project is conceived in its lexicon, its syntax, and use of materials, in systematic opposition to context, and not only in opposition to a uniform context characteristic of the pre-industrial past but also to the less uniform context of the present. How do we propose to counter the practice of disassociation that rules the growth of fabric and urban systems? We will need a clear awareness
the fourth typology of the proper role of reading and its specific ends. The entomologist only studies ants; the architect “makes them”. Readings are therefore biased insofar as they are meant not only to determine facts but also ferret out principles for design through the analysis of processes. If the design is correctly understood in terms of critical consciousness it must also try to put together the pieces of a spontaneous consciousness with positive pieces of the Modern Movement, particularly the research dedicated to social housing. In the operative phase the typological process lends itself to interesting considerations that can be articulated on three levels, depending on whether it refers to the restoration of a building, the reorganization of a fabric, or an ex-novo project. Regarding the problem of restoration58 we note that today’s monuments are presented with a limited number of changes and are recognizable in the organism by stylistically identifiable criteria. In addition, they often come with archival documentation that includes the names of both the client and the architect. The critical restoration of a monument is limited to the consolidation of the static structure, the elimination of distributive inconsistencies and superfluous additions; its aim is adaptive re-use. This is not the case for residential buildings. As in the contrast between spoken and written language, the experience represented by the construction, modification, and adaptation of dwellings is far more numerous but rarely documented. If a built object endures longer than a human lifespan and many houses disappear only to be replaced by others, permanence continues to influence the structure by obliging it to retain characteristics such as size, position, and conformity with the shape of the lot. When restoring a residence it is rare to encounter what is obviously an original structure, outside the bounds of an archaeological excavation. Often minimal changes guided by new revised concepts of the house bring the building up to date, and if it is still being lived in, it should, for all practical purposes, be thought of as a contemporary building. The skinning of the walls that anticipates the scientific restoration of many houses of the Moorish quarter of Albajcin in Granada reveals the Nazari origins of the houses, the openings, occlusions, changes in purpose and distribution in the succeeding Mozarabic era, the variations in height, the destruction of parts.59 Two typological processes for residential types can be illustrated by Essaouira/Mogador on the Atlantic shores of Morocco, and by Cairo. In the approximately two hundred years since the founding of Essaouira, a relatively simple process has taken root: the first building foundation in the casbah quarter near the sultan’s palace is essentially the basic courtyard house common to the rural Atlas region. This small (10 m. across) singlefamily dwelling features a main room built on the north side and open on the south. Successive variations show vertical growth of up to four stories, and in recent times the structure has been host to many families rather than a single one. As the city prospered in the mid-nineteenth century the leading type was the two-story house with rooms symmetrically arranged around a central courtyard. This in turn was organically joined to a regular urban building tissue perforated by a network of broad streets.
49
50
after amnesia
This type left its mark on the new zones of La Lagune and Skala planned in the seventies in two modified forms. In La Lagune, a three-story type is enclosed on three sides. Its height varies, and a small courtyard is covered, thus creating serious hygiene problems. The type found in the Skala area is similar but its front is half the size and the light well is on the long side. It appears that even the most current building type planned by Essaouira’s city engineers is not very different from the matrix type found in the casbah. But the case of Essaouira is interesting in the theoretical turns it takes due to its mellah, home to Jews since 1947. The quarter was originally built up very quickly at the end of the nineteenth century owing to a population explosion.60 This led to several synchronic variations in the leading type in fairly regular building tissues located primarily at the edge of main transverse routes and along the perimeter of the city walls. Further increments led to the completion of the urban fabric with infill building tissues featuring poorer and less advanced types. The 1947 large scale emigration from the countryside immediately necessitated a conversion to multi-family houses, the filling in of open spaces, and dizzying vertical additions. These processes compromised the equilibrium and the functionality of the buildings. Cairo, on the other hand, underwent a different process of conversion. The medieval single-family row-house gave way to the apartment house as a leading type through a long process involving the combining of plots and the addition of extra stories. In both Essaouira and Cairo, it is clear that in restoration the primary concern was whether the type could bear the weight of the required
Fig. 19 Typological process of dwellings in Essaouira, Morocco. Note the constant surface growth and the ever-increasing search for symmetry in the evolution of the new leading types until the second half of the nineteenth century. The model imported during the colonial period is a linear combination of four rows and generates a local variant generally used for apartment buildings.
the fourth typology changes. Do the width of the old streets and the distance between the houses adversely affect the inhabitability of the house under restoration? Are the building type and the building tissue still coherent? To carry out restoration without paying attention to pre-existing conditions means risking an untenable situation susceptible to decay. A precise reconstruction of the typological process would avoid a situation in which the reduction of a space’s volume is indiscriminately based on aesthetic criteria, inspired by methodologies developed in the restoration of architectural monuments. Suffice it to say that the restoration of a building should always be a synchronic variation of an earlier leading type in order to make it consistent with present-day life. In the event of manifest inconsistencies with the building tissue, a step backward with respect to the fitting of the current building is justified to update the type, provided that it is compatible with the parallel process in that location and therefore with the imposition of the building tissue. In other words, it is possible to operate on incoherent elements, such as superfluous additions that open up spaces, super-elevations, garages etc., when multi-family subdivisions are illegitimate, because the first building and tissue cause an incompatibility between volume, number of stories, and number of people and the order of streets and inherited property divisions. In other areas the building tissue may tolerate unauthorized volumetric additions. One thing is certain: the restoration of houses cannot refer a-priori to a single leading type for the whole city, but must incorporate a range of synchronic variations consistent with different tissues in different areas. With reference to the reorganization of building tissues we note that reestablishing continuity in tissues traumatized by war, natural disasters, or uncontrolled building speculation is important. Contrary to the tendencies of many city planners, in the “mineral” city of stone and brick it is better to re-establish the continuity of the fabric than to open up spaces in a misguided attempt to improve the quality of urban green spaces or to provide parking. The tragic events of Bosnia-Herzegovina push the world’s conscience toward an accelerated reconstruction of cities, to close an open wound as soon as possible, especially in the feelings of the inhabitants and then in the body of the city. Two opposing attitudes generally prevail in similar predicaments: one, driven by popular nostalgia, wants to rebuild everything as it was and where it was, extending the concept of restoration to the entire body of the city. This public opinion, in conserving the image of the past city, is often content with a facade image, as in Warsaw after 1945, and the problem of the authenticity of what is rebuilt is never considered. The other attitude sees in the great destruction and the vast empty spaces left by demolitions a unique chance to design the utopia of the future city. These are generally architects who wish to form the city in their own image, fill in the holes with fantasy architecture that is rarely in keeping with the context, but instead with the abstract, personal and self-referential poetics of architecture journals, paying no heed to the inhabitants. It is the same type of architecture of the last fifty years, an expression of the Modern Movement in Bosnia-Herzegovina that had tried mightily to break with history, in the name of the invention of utopia.
51
52
after amnesia
I believe that a third position is more urgent and important: the reconstruction of the cities of Bosnia-Herzegovina (but also in Lebanon and in Palestine) can not be only a problem of (sacrosanct) indemnity to leave in the hands of technicians, but the chance above all for a radical rethinking of the meaning of the contemporary city and a deep critique of the modern city. Mostar, which lost eighty percent of its buildings in the central part of the city, can become an experimental laboratory for a change of direction of the typological processes, capable of sending out a strong message even beyond the Bosnian border. All cities are the product of the slow sedimentation of the built environment that happens in phases, in which the next one has absorbed and retained as much as possible from the former, which acted as its reference and model. In the building history of cities in the pre-modern era, building types and types of fabrics and construction techniques evolved and were perfected through a slow process that has always allowed for the metabolizing of the preceding experience according to the principle of the most yield, which always presides over building operations when the available technologies are limited. Ottoman culture designed the Bosnian landscape with respect to the preceding local tradition and the multicultural reality with quarters (mohalla), in which a wise dialectic is
Fig. 20 An academic project of infill in the urban fabric of Aleppo. Thesis project of G. Ambriola, A. Dimasi, P. Lestingi, V. Lorenzini, G. Minervini, G. Roppo. Adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari.
the fourth typology
53
Fig. 21 Mostar. Academic project of a block along the boulevard. The existing building, from the 1970s, that represents effectively, with its disarticulated and detached plan of the city streets, the disruption of the urban tissue, has been treated as a “natural obstacle.” The building type of the linear nineteenthcentury block with porticoes is laid out along the avenue, while the building type of the courtyard house is inside the block. Thesis project of F. Basile, E. Delia, L. Di Bari, L. Lionetti, M. Loisi, A. Margiotta, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari.
established between monument and residential building. The former, even if they refer to the canonical types of mosques and külliye codified by the Sublime Gate, never reach the enormous scale, magnificence and rhetoric of the monuments of the capital. The second of the common courtyard type with sofa comprises a continuous fabric of high quality. The Austrian administration installed in Bosnia-Herzegovina after the Berlin Congress, even if importing Middle-European construction techniques and building types, has carried out the expansion of the new city in homogeneous and continuous parts of the fabric, based on the closed block; even more importantly, it has accepted a dialogue with the preceding Turkish city in the continuity of use, in the types and even in terms of style. The most evident cases are the building completion of the main street and the square of Musalla. From the political standpoint, the attempt on the part of the Austrian administration was focused on the
54
after amnesia
one hand to resolve the problems of the city with rational engineering interventions, but on the other hand, it was focusing on underlining with the choices of architecture the dialogue of the different cultures, under the protective and benevolent eye of Emperor Franz Joseph. The Modern Movement willingly detached itself, building a city made up of isolated building episodes in contrast to one another and with scarce adhesion to the spirit of the place, breaking the continuity of the typological process and deviating from it, creating places of cultural segregation, a premise of civil dispute. The reconstruction of Mostar must become the opportunity for a change of course, taking advantage of the fact that a great part of recent building has been knocked down by demolitions and only very few buildings are still standing as relics on the water’s edge. The rudder can be straightened by re-evaluating the continuous city and giving meaning to the typological processes that occurred before the break of the Modern Movement. This is not a case of rebuilding as it was and where it was, but one of rebuilding as if…: as if the Austrian city of closed blocks were not broken up and diluted into the modern city, as if the Turkish city had not disappeared from the blows of the Socialist Society, as if the surviving modernist buildings could be included as “ruins” in the new continuous city. Constantine in Algeria is a good example of a contemporary city administration in a quandary over its historical fabric. The city is built on a rocky plateau surrounded on three sides by a canyon which acts as a natural fortress. Its growth since Roman times has been organic and has alternated between loose planning and spontaneous adjustments. When the French colonized Algeria, Constantine possessed a continuous urban fabric of courtyard houses, and not even the French colonial intervention of two roads as a cross-axis at the center of town radiating from the Piazza XX November could upset it.61 The incisions, in fact, followed the fibers of the building tissue and were integrated with building types whose internal open spaces, or backyards, were not very different from the indigenous courtyard house. But following the independence of Algeria, two sizable areas near the cliff and inhabited by locals rapidly deteriorated, were abandoned, and finally demolished. Today two large dusty patches mark the space where houses, mosques, and lively bazaars once stood. Algeria’s current economic crisis has at least spared its historic town centre from being subjected to a common solution: the building of linear, parallel, multi-storied prefabricated blocks oriented on the helio-thermal axis. But the key question remains: Where to go from here? Constantine requires urban restoration aimed at the re-composition of the environmental unity of its genius loci, the suture of the wounds and the refastening of the preceding routes, not to mention the recovery of the common types from the phase of the most coherent past. The design of such a project should, however, not simply be a wholesale reinstatement of what was there before. Such a solution is acceptable only in the case of a meaningful monument or collection of monuments such as the old city of Warsaw or the Aqsa Mosque following
Fig. 22 Academic project for an arts-and-crafts centre in Galata, Istanbul. Thesis project of Patrizia Sicuro, advisers L. Micara and A. Petruccioli, Università di Roma “La Sapienza.”
the fourth typology
its devastation by fire. In most cases, design should be based on a careful reading of the typological processes of the area that best describe the typological expression that preceded the catastrophe in order to recapture the process from that instant. This would require using a synchronic variation of a fabric for reconstructive purposes and not a leading-type fabric such as is usually found in the outskirts of a city. The final product is certain to differ with respect to the original because it uses the synchronic variation most closely linked to the original building project. Conceptually the operation is not very different from the reconstruction of a single building. The choice of variant type to introduce into the fabric can be perplexing. Those who tend toward the picturesque often choose types too close to an earlier leading building type. A plan to refill a fabric in the Galata area of Istanbul (see fig. 22) where many buildings were missing illustrates this dilemma. An “Ottoman type” would be incongruous in lexicon, syntax, and materials with the surrounding nineteenth-century apartment houses. If this project were carried out over time and following the principle of urban self-regulation it would inevitably be accomplished by further demolition and by drastically increasing volume because people will try to approximate the recent leading type. Another uncertainty results from the amount of time that elapses
55
Fig. 23 Academic project for the peripheral areas of the medina of Meknès, Morocco. Thesis project of Gaetano Arcuri and Anna Di Pasquale, advisers L. Micara and A. Petruccioli, Università di Roma “La Sapienza.” Note the courtyard tissue developing in the voids at the bottom between the low-income residential slabs.
56
after amnesia
between demolition and reconstruction: the use of an empty space (often nodal with respect to the city) as a parking lot, soccer field, or the more socially complex public square determines the new process. At the moment of redesigning the question becomes whether to pursue a new typological process with new building types and contemporary fabric, perhaps even paradoxically imposed on new streets, or to use a reconstructive synchronic variant. In this case, it is the community rather than the designer who is responsible for the decision. Finally an ex-novo project in an expansion area involves so many considerations that it is both a far-reaching and significant design exercise. Considering that today, as opposed to perhaps the desert, a virgin territory in which to exercise creativity ad libitum does not exist. Even an empty territory in reality reveals traces of routes, productive activities and sparse settlements that are expressions of the stratification of history, and not so differently the archaeological finds buried beneath the ground. All these signs, instead of being detrimental bonds to the project, are occasions with which it must be put in tune. In synthesis, an ex-novo project ripped from its context does not exist and an important corollary can be deduced from it: that beyond specific operative instruments, there is no conceptual difference between a restoration project and an ex-novo project. Today in residential typological processes population mobility and
Fig. 24 Academic project of a mat fabric in the periphery of Essaouira. Like a fusing liquid, the tissue fills the empty areas, which act as “moulds.” The logic of the urban tissue in the periphery is overturned: instead of small isolated houses surrounded by insignificant leftover spaces, a continuum of interior spaces hierarchized by the public, to the semi-public to the intimate space of the house. It is possible to propose the logic of the traditional city for the expansion of the new cities. (Thesis project of A. Cutrone, P. Genchi, Gm. Lozupone, L. Massarelli, A. Moccia, D. Panaro, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari).
the fourth typology
modern communications have encouraged diatopic processes (which were secondary in the pre-industrial city). For example, the influence of television allows an economically dominant nation like the United States to impose across the ocean the “ranch style” house made famous by soap operas as a leading type ( it is actually a model), and a cultural leader in the Arab world like Egypt to impose the Egyptian house or fasha to the Maghreb as a model admired for its Pharaonic proportions. A rupture in the homogeneity of traditional social structures and the emergence of multi-cultural societies has led to a fragmentation of urban unity. The American folklorist H. Glassie remarked on the paradox that as diverse groups assert their cultural individuality and people proclaim their rights as individuals, building responds by stubbornly clinging to a single universal, generic leading type.62 Thus today the role of the designer inevitably increases, but this does not necessarily mean that the only option given to a multi-cultural, non-
57
Fig. 25 A piece of the preceding tissue shows a pair of neighborhood units organized around semipublic courtyards of equal dimensions. Each house has its own small courtyard on different levels for air and light and sheltered for introspection. (Thesis project of A. Cutrone, P. Genchi, Gm. Lozupone, L. Massarelli, A. Moccia, D. Panaro, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari).
58
after amnesia
homogeneous, and critically conscious society is to abandon design to the anarchy of mercantile choices disguised as aesthetic preferences, especially on the city scale. In the real world, the true victim of alienation is not so much the house type -it is still firmly anchored in the European apartment building and the American single-family detached house- as it is the fabric, exiled to the periphery of European cities and the suburbs of American ones. The fabric is an endangered species, squeezed on one side by the multiplicity of types pushed into use beyond their limits,63 and on the other, by a kind of urban planning that is opposed to the homogeneous historic fabric in the name of the myth of the machine city. It is useful to have a critical knowledge of typological processes and their place in the initial phase of a design as a reference point for action. Rather than prescribe, as a normative body does, typology serves to orient by revealing tendencies without eliminating the synthetic intuitive responsibility involved in making good design choices.
Fig. 26 Academic project of houses for temporary residence in the bazaar of Antakya. The project brings about a synthesis between the plastic-mural tradition of Syrian derivation in the heavy base, and the Ottoman elasticwood tradition of the attic. Compare with the house in fig. 61. Thesis project of Stella Bellomo, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari.
the fourth typology A good example is found in the work of the Amsterdam School, led by H.P. Berlage and younger architects such as M. De Klerk, P. Kramers, H. T. Wijdeveld and M. Kropholler-Staal. In the quarter of AmsterdamZuid their work provides an excellent example of the fusion between typological process and creativity. The designs for the subsidized building constructed in the inter-war period on the outskirts of Amsterdam were an architectural realization of official expansion plans designed by Berlage and city planners. In Amsterdam, architectural imagination freed from anarchy defines the envelope of the building that is nevertheless based on the cultural traditions and trends that produced the apartment house and directly linked through typology to the city’s typical middle-class house. Amsterdam is the best example of how a stable bourgeois society is able to produce shared types so rooted in the social tissue that they are able to absorb the eccentricities of Expressionist architects. In these pages we have tried to describe a method of interpreting built reality that would be objective and could serve as a guideline in the design process. Moving from the definition of structure by Piaget and recognizing in type the same quality of totality, transformation, and self-regulation, we arrived at an understanding of the meaning of type as a synthesis a priori. The a-priori type pre-exists and anticipates any action of building; it is characteristic of the period of spontaneous awareness, but in a period of crisis similar to the one we are living in today, it is reduced to fragments. It can only be recovered a-posteriori by a rigorous analysis supported by a reflective awareness. A further point, implicit in the association of type with history, is the introduction of the factor of time in typology. The idea of process is the key factor that qualifies our method and distinguishes it from its predecessors. To introduce the concept of time means to introduce a dynamic factor in historic transformation over time. It could be no different since type is the expression of a society in a circumscribed area and in a given time. It is necessary to accept the complexity of architecture in the infinite relations established in history between residential buildings and special buildings. It will offer new and interesting horizons in the field of renovation, urban preservation, and design. Even though the built environment is global, to simplify the process of reading in the next chapters we will proceed from the simplest level to the most complex, from building to tissue, to fabric, to city, and finally to territory.
59
60
after amnesia
notes 1. N. Pevsner, An Outline of European Architecture (Harmondsworth, 1960), p. 1. 2. N. Pevsner, A History of Building Types (Princeton, N.J., 1976). 3. “Building” in Webster’s New Collegiate Dictionary, ( London,1980 ) 4. This is my definition of building, and in the following pages the term “building” will be used according to this definition. 5. Karl Scheffler, Introduction to Heinrich Tessenow’s Die Architektur der Groszstadt (Berlin, 1913), translated into English as Housebuilding and Such Things (Cambridge, Mass., 1989). 6. L. Quaroni, Progettare un edificio. Otto lezioni di architettura (Milan, 1977). 7. G. Giovannoni, Vecchie città edilizia nuova (Turin, 1931). Not until the 1950s did Italians renew their discussion of building by arguing about minor versus major or popular versus refined architecture. The discussion was framed by Croce’s famous distinction between poetry and literature and was ill-suited to the architectural debate because building must be considered from both the social and the logical points of view. See R. Pane, Attualità dell’ambiente antico (Florence, 1967). A brief work by G. Pagano Pogatschnig, Architettura rurale italiana (Milan, 1936), is also intriguing for the interest it generated in the vernacular architecture of the Italian Rationalist Movement. 8. There is a vast literature on the nineteenth-century bourgeois city. An attempt at revision that places emphasis on social values, livability and building quality is: A. Petruccioli ed., Rethinking the XIX Century City, AKPIA, Cambridge, MA, 1999. 9. André Raymond questions the Orientalist’s prejudice in describing pre-colonial Arab cities as being prey to economic recession. On the major Eastern cities of the Ottoman Empire, see A. Raymond, The Great Arab Cities in the XVIXVIII Centuries: An Introduction (New York, 1984). Although the material available on the Ottoman period is limited, Raymond is able to cite several indices that indicate steady growth in cities such as Aleppo and Cairo. See, for example, A. Raymond, “Le déplacement des tanneries à Alep, au
Caire et à Tunis à l’époque ottomane: un ‘indicateur’ de croissance urbaine,” in J. B. Pascual, ed., “Villes au Levant”, special issue of Revue du Monde Musulman et de la Méditerranée, 55-56 (1990), p. 9. 10. I am not referring to arabisances, or the sincere if somewhat clumsy attempt to integrate Arab culture into Western culture through the use of stylistic elements taken tout court from local architecture and used as decorative items. Instead, I refer to instances of structural continuity such as the adaptation of types, measures, and construction standards introduced in colonial techniques, and to the use of local materials, etc. On arabisances, see the special issue of Lotus International 26 (1980) for essays by J. Gubler, A. Barey, G. Baudez, etc.; see also F. Beguin, Arabisances, Decor architectural et trace urbaine en Afrique du nord 1830-1950 (Paris, 1983) 11. J. Piaget, Le structuralisme (Paris, 1968), pp. 6-7. On the analogy between structure and organism see : G. Strappa. Unità dell’organismo architettonico. Note sulla formazione e trasformazione degli edifici ( Bari, 1995). 12. From “I Love Beginnings,” lecture by Louis Kahn at the Aspen Design Conference, 1973. “If you ask a brick what it wants to be, it will answer: I want to be an arch…This means understanding order.” 13. Too much emphasis on utilitas can result in a building in which circulation is more important than space. Too great an emphasis on firmitas leads to the hyperbolic use of technological components which can degenerate into an overuse of high-tech ornamentation. Finally, too much venustas can result in a dramatic and expressionistic stress on empty form. 14. F. De Saussure, Course in General Linguistics (London, 1974). 15. L. Hjelmslev, Language: An Introduction (Madison, Wisconsin, 1970). 16. The city gates allow inhabitants to leave the city boundaries and at the same time prevent undesirables from entering. They may be augmented by a protective external gate meant to discourage potential invaders, but this message is intentionally obscure and therefore a non-message. The gates could also be thought of as a synecdoche, a rhetorical device that represents a whole by one of its parts. If elaborately decorated
they can also make a statement to the outside world about the economic power of the city. But whatever else they are, the gates are first and foremost an opening in the continuity of the wall. Gates must maintain the integrity of the walls, ensure protection, and hold themselves up. The arch that is often seen as a decorative element of the gate, and therefore just a symbol, is in reality the logical solution for transferring the weight above to the walls on either side. 17. A good example is the polylobated arch which nowadays has become synonymous with Islamic culture. Today it is usually made on site out of concrete (one Belgian firm produces them in fiberglass), and is planned for a particular room which makes it structurally superfluous. 18. G. C. Argan, Modulo-misura e modulooggetto, Progetto e destino (Milan, 1965), p. 113. On the same topic, see A. Petruccioli, “Modulo,” in L. Quaroni , ed., Progettare un edificio. Otto lezioni di architettura, (Milan, 1977), pp. 173-176. 19. G. De Carlo, “Note sulla incontinente ascesa della tipologia,” Casabella, Jan.-Feb. 1985: 50. “Besides being confirmed by solid normative tools and imposed by the convenience of production, rigid types - or stereotypes - have also become repressive in the sense they do not allow variations, additions, subtractions. In other words they do not permit the participation of those who will use them.” 20. G. De Carlo. “L’architettura della partecipazione”, in L’architettura degli anni settanta (Milan, 1971), p. 105. 21. “While the mason placed brick upon brick and stone upon stone, the carpenter took his place alongside him. Happy are the sounds of the axe striking blow after blow. He builds the roof. What kind of roof? A beautiful roof or an ugly roof? He doesn’t know. The roof.” (A. Loos, “Architektur”, Trotzdem (Vienna, 1982), pp. 90-91. 22. “To assume homogeneity means developing a noteworthy sensitivity, a sensitivity in only one direction; because it is necessary to concentrate all one’s attention on every subtle detail; this is the limitation which characterizes uniformity and also order.” Tessenow, House Building and Such Things, p. 17. 23. Carlo Enrico Rava is perhaps the least
61
known, but decidedly the most important figure as a theorist of the Rationalist Movement (he was leader of the Gruppo 7 in 1926) and a participant in the debate on modern colonial architecture. His ideas on colonial architecture begin with the rejection of Moorish (and other) styles as inspirations for Libyan architecture, which he claims is of classical Roman derivation. Ultimately he sees “an extremely fine Mediterranean intonation that clearly relates it to all other architectures of southern origin” (C. E. Rava, “Una architettura coloniale moderna mediterranea,” Domus ( 1931 ), 39. On the same subject see also idem, “Costruire in colonia,” Domus (August-October 1936), 28-30; and idem, Nove anni di architettura vissuta 1926 IV-1935 XIII (Rome, 1935 ), 103 ff. 24.S. Muratori, “I Caratteri degli edifici nello studio dell’architettura,” Inaugural lecture, Course on the Characteristics of Buildings, IUAV,Venice, 1950, p. 15. 25.Ibid., Muratori further adds, “What is more, the most orthodox application of that idealist thought which seeks to negate type does not suffice to deny the existence of true collective expressive creations - manifest in some typical spatial and structural intuitions - which make up the architectural core of an entire civilization.” 26.“Types are shared properties within a culture. Everyone - builder, designer, user, is familiar with them.Yet types such as the Venetian Gothic palace, the Amsterdam renaissance townhouse, the Georgian terraced house, or the Pompeian courtyard house were never formally described by those who made and used them. Types only exist in a social body.” J. Habraken, “The Control of Complexity,” Places 4, n. 2 (1987), p. 7. 27.“A type, in the sense we intend, is neither a general category like ‘church’ nor does it consist only in a particular instance, like Richardson’s Trinity Church. In our sense, types should be seen as particulars that function in a general way, or as general categories that have the fullness of particulars. Examples might be the New England green, Oxbridge lawns, pavilion, and cave mentioned in William Porter’s essay” (D. Schon, “Designing: Rules, Types and Worlds,” Design Studies 9, no. 3 (July 1988). 28.A. Petruccioli, “Alger 1830-1930. Pour une lecture typologique des immeubles
d’habitation,” Environmental Design: Journal of the Islamic Environmental Design Research Center (1992), pp. 104-117.
1993), p. 322.
29.I. Calvino, Invisible Cities (Orlando, Florida, 1974), p. 139. The cataloguer harbours the illusion of possessing the world as in Kublai Khan’s atlas: “The catalogue of forms is endless: until every shape has found its city, new cities will continue to be born. When the forms exhaust their variety and come apart, the end of cities begins. In the last pages of the atlas there is an outpouring of networks without beginning or end, cities in the shape of Los Angeles, in the shape of Kyoto-Osaka, without shape.”
36.For a more detailed explanation, see P. Maretto, Realtà naturale e realtà costruita (Florence, 1980).
30.S. Muratori, “Saggi di metodo nell’impostazione dello studio dell’architettura,” Rome, 1946. 31.From asymptote, a line which appears nearer and nearer to a curve but does not meet it within a finite distance. 32.A point must be made here: the building, the city or the territory are not written texts. This misunderstanding led in the 1970s to the abuse of semiotic categories and to the attempt to assimilate architecture to the communications sciences, with quite sterile results, so that the centrality of the discipline was no longer evident. Communication and significant phenomena are two different things: a “do not enter” sign conveys a clear and univocal message; the message conveyed by the gates of the city, for example, is ambiguous, stratified and has many meanings. 33.G. Caniggia and G.L. Maffei. Composizione architettonica e tipologia edilizia. Lettura dell’edilizia di base ( Venice, 1982), p. 95 34.Karl Kropf wrote a monumental doctoral thesis comparing the reading methods of Gianfranco Caniggia and M.G.R. Conzen. On the subject of the scalar complexity between building materials and city he proposes a nine-level hierarchy: building materials; structural elements; rooms; buildings; plots; plot series; tissues; combinations of tissues; and combinations of objects on the previous level which he refers to by the following Latin terms: materia; statio; tectum; aedes; fines; sertum; textus; sedes; complures. See K. S. Kropf, An Enquiry into the Definition of Built Form in Urban Morphology (Birmingham,
35.Ibid., p. 175.
37.B. Croce, Logica come scienza del concetto (Bari, 1908). 38.Popper postulates a third entity or world in which the objective facts of thought have a life of their own, and this has interesting implications for the theory of a project. See K. R. Popper, Unended Quest, an Intellectual Autobiography (LaSalle, Ill., 1976). 39.This topic is dealt with in G. Marinucci , ed., Saverio Muratori: Metodologia del sistema realtà-autocoscienza (Rome, 1978), a publication of his last lectures in 1972-73. 40.Muratori’s theory and methods are found in S. Muratori, Civiltà e territorio (Rome, 1967). This work contains a detailed version of the table of elements. 41.See G. Cataldi, All’origine dell’Abitare (Florence, 1986), and “La capanna di pietra. Ipotesi evolutive di inquadramento sistematico,”in G. Cataldi, ed., Le ragioni dell’abitare (Florence, 1988), pp. 79-90. 42.S. Muratori, Autocoscienza e realtà nella storia degli ecumeni civili (Rome, 1976), p. 119. “Chinese cities were very large, indeed quantitative, and organized like the seriality of their territory: they were divided into large quarters, cities within cities and enclosed by walls for the purpose of greater control.” 43.In using these categories it should be made clear that calling something organic or serial does not constitute a value judgment of a culture, but only an evaluation of a precise historical area. 44.A. Petruccioli, Fatehpur Sikri (Berlin,1992). On Bukhara: A. Petruccioli, ed., Bukhara.The Myth and the Architecture (Cambridge, MA, 1999), and H. Gaube, A. Gangler, A. Petruccioli, Bukhara.The Eastern Dome of Islam (Stuttgart 2004). 45.G. Caniggia and G. L. Maffei, Composizione … la lettura, p. 72. G. Cataldi, F. Farneti, R. Larco, F. Pellegrino, P. Tamburini, Tipologie primitive. I “tipi radici” (Florence, 1982); G. Cataldi, “Ipotesi di
62
after amnesia
46.G. Gurvitch, Determinismes sociaux et liberté humaine (Paris, 1955), pp. 38-40.
the sixteenth-century transformation of the apartments belonging to Europeans which were found on the upper stories of khans and caravanserais. See J. C. David and D. Hubert, “Maisons et immeubles du debut du XXe siècle à Alep.” Les cahiers de la recherche architecturale, 10-11 (April 1982), p. 105.
47.G. Kubler, The Shape of Time: Remarks on the History of Things (New Haven, Conn., 1962), p. 14.
54. A. Petruccioli, “Alger 1830-1930,” in A. Petruccioli, ed., Algérie: Mémoire et architecture, Environmental Design 1-2 (1992), p. 104.
48.The term diachronic comes from the Greek dia (through) and chronos (time); syntopic from syn (with) and topos (place or site), literally, in the same place. Diachronic can also refer to a bridge between two sets of time, and syntopic can mean following the rhythm of the site. The idea of process includes diachronic motion, but process need not always be confined to the same site, although the first step in analysis can generally assume this to be the case. It is obvious that if we study a process at a site we start with all the mutations occurring at that site and only then do we extend the analysis to other cities or environments to enrich our understanding of the original process.
55.Caniggia and Maffei, Il progetto dell’edilizia di base, 223 ff.
inquadramento metodologico per uno studio sistematico degli sviluppi diffusivi della abitazione umana” in G. Cataldi, ed., All’origine dell’abitare, (Florence, 1988), pp. 25-36
49.Bernoulli provides a penetrating look at the permanence of real estate, even though he concludes that private property and its fragmentation are the principal ills of the modern city and predicts a collectivization of urban land. See H. Bernoulli, Die Stadt und ihr Boden (Erlenbach-Zurich, 1946). 50. For a discussion of Back Bay urban planning, see S. Stenti, “Boston Back Bay: Atene d’America,” Storia della città, 49 (1990), pp. 7-42; W. Whitehill, Boston: A Topographical History (Cambridge, Mass. 1959). 51.From the Greek dia, through, and topos, site; something that happens in different areas or places; something that does not belong to a place. Because of the meaning given to type, I should have used in the text the term model, as something alien that is copied acritically. 52.G. Caniggia and G. L. Maffei, Composizione architettonica … edilizia di base, pp. 76 ff. 53.Jean Claude David traces the birth of the sala passante crossing hall type used extensively in the Aziziye quarter to the local typological process, starting with
56.Kublai asks Marco, “When you return to the West, will you repeat to your people the same tales you tell me?” “I speak and speak,” Marco says, “but the listener retains only the words he is expecting” (Calvino, Invisible Cities, p. 135). The idea that intentions influence the selection of data is shared by Pierce and Feyerabend (see P. P. Wiene, ed., C. S. Pierce, Selected Writings, [New York, 1958]; and P. Feyerabend, Against Method [London, 1975]). 57.“ Like ingenuous functionalism, such theories consider construction to coincide with the conclusive process of design, with the pure material act of building, and not with the options and strategic choices that lead to an accomplished work……One could affirm that if a really “creative” act exists in the project, it is constituted by the first constructional choice….. Construction cannot be reduced to the simple act of building or practical execution of a project, but must be considered a “high” synthesis of homogeneous cultural areas which through technique ( in the old sense of techné ), express specific values. From Claudio D’amato Guerrieri “The Contemporary Architectural Project”, in Area , 39 ( 1998), pp. 4-7 58.The concept of restoration has undergone a profound change in Italy thanks to the Marconi school. The separation of modern materials from the traditional materials of the monument as an obsession for the authenticity of the material is at the heart of the Charter of Venice (affiliated from the start with Modernism’s principle of discontinuity). Paolo Marconi recognizes the failure of elastic materials in restoration and subscribes to the idea of the maintenance and continuity of materials. His argument is primarily based on the acknowledgment that traditional materials are more the
victims of cultural marginalization than of genuine and obvious obsolescence. Whether materials such as brick or wood in all their forms are obsolete remains to be proved. See Marconi. P. Il restauro e l’architetto.Teoria e pratica in due secoli di dibattito (Venice, 1993). In the 1970s the old centre of the city of Bologna was subjected to the first scientific attempt at historical preservation. The results stirred up a good deal of interest and were published in P. L. Cervellati, R. Scannavini and C. De Angelis, La nuova cultura della città (Milan, 1977). The building tissue and types did not emerge unscathed from this pioneering effort: the typological reconstruction used too few leading types and the end result is highly schematic. For an elaborated version of this theme with particular reference to the concepts of revival and reuse, see A.Petruccioli, “Alice’s Dilemma,” in Typological Process and Design Theory (Cambridge, AKPIA, 1997). 59.The renewed attention of Spanish culture to its own Arab past and the high competence of research institutions like the Centro de Estudio Arabes of Granada favor the scientific restoration of basic architecture as well. On the Arab architecture of Toledo, see: Jean Passini, Casas y casas principales urbanas. El espacio domestico a fines de la Edad Media, (Toledo, 2004). With regard to Granada, we cite the passionate work of Antonio Almagro and Julio Navarro Palazon. In Albaicin during the Mozarabic era and during the subsequent Baroque period, the Nazari character of the houses was overturned. In particular, the important sitting room (diwan) with a southern exposure developed a height and a half between five and six meters with a wood covering in two layers. During successive additions the height of the sitting room was reduced and the roof torn down. Only the removal of the plaster permits the finding of the locations of the joists and the position of the old roof. 60.D. J. Schroeter, Merchants of Essaouira. Urban Society and Imperialism in Southwestern Morocco, 1844-1886 (Cambridge, Mass., 1988); also, H. Froidevaux, “Une description de Mogador en 1765,” Annales de géographie 2 (1893), pp. 394-398. 61.B. Pagand, “De la ville arabe à la ville europeenne: architecture et formation urbaine à Constantine au XIX siècle,” Figures de l’orientalisme en architecture, special issue of Revue de l’Occident
notes
Musulman et de la Méditerranée, 73-74 (1994), pp. 281-294. 62.“In the Anglo-American world of the eighteenth century the farther people moved apart, the more alike their architectural expressions became. Through time, the increase in individualism was accompanied by a contraction of the culture: the dominance of fewer house types, less variation within types once the fully symmetrical design had been achieved, and a diminishing of the inventory of detail and decoration. Houses once red, yellow, and white became white” (H. Glassie, Folk Housing in Middle Virginia (Knoxville, Tenn., 1972), p. 182. 63. The most glaring example of this is found in Rome in the working-class quarter of Corviale, built at the end of the late 1970s. The synthesis between the Corbusian scheme of gratteterre and the enormity of the container is realized in a building one kilometer long and fourteen stories high which straddles the hills south of the city and whose volume has changed the climate.The impact on the undulating landscape of the hills and on the social fabric of the inhabitants has been disastrous.
63
64
after amnesia
The rab of Kait Bey in Cairo. Monumental entrance (from P. Coste)
65
the building type If we analyze the construction process of a house, we see that the builder possesses in thought the form of the house; he knows what the fact of being a house is. In a certain sense the house has its beginnings in house: in something immaterial (its concept) that generates something that includes the material. Aristotle, Metaphysics A building - the first level on the hierarchy of scale - can be looked at in two ways: it can be broken into components, or it can be read according to a progressively more complex hierarchy of levels - elements, structures of elements, systems of structures, and finally organism of systems. It is best to begin at the smallest scale for two reasons: first, the architectural organism is most familiar to architects and students alike; it is always broken down, even if only unconsciously, in the course of a project or in verifying structural and metric calculations. Second, the largest scale or territory which includes all smaller scales is too complex because each progression in scale increases the specialization of the components and the hierarchization of the parts according to a quadratic law. It is therefore wiser to read the built environment beginning at the simplest and the most concrete level and save the reverse process - from territory to building - as a means of verification. Though for the sake of simplicity in this chapter we will deal with the building out of context, it is well to bear in mind that not only is an organism composed of elements, but it in turn forms an element in a larger-scale organism. The hierarchy of relationships functions so that the farther apart the elements, the less closely they are related. The reading of an organism at any scale is informed and reinforced by the readings at the preceding smaller scale as well as the reading at the succeeding larger scale. The analogy with photography is once again useful here: rather than read the reality with a fixed focal-length lens or change lenses, we can use a zoom lens to capture the seamless continuity of the scalar changes. Lastly, we attempt to identify the different branches of the “Islamic” typological processes that belong to the Mediterranean tree, dealing with the house of the east with its Ottoman developments, the Venetian house and the special case of the Cairo house. The problem of the break in continuity of the typological process and its successive revival is examined in the last section of this chapter, dedicated to the colonial house as exemplified in Algiers. A problem in structuring this book is that it runs the risk of becoming an encyclopaedic documentation of the typological processes of the Mediterranean basin, an accumulation of data lacking any synthesis. The result would be to lose sight of the original goal, which was to establish principles in a structured framework that would guide the reading of a built environment. One solution might have been to analyse synchronically a single area, a state such as Tunisia, and draw guiding principles from
66
after amnesia
it. However, such a study would not elaborate the whole spectrum of typological theory, as no single case can encompass all possible issues - thus the choice of a synthetic platform that would use a variety of cases in the Mediterranean basin for comparative analysis. Fig. 27 Diagram of the variants of the elementary cell, located according to the topography and access to dwellings. The most important variants can be summarized in three main categories of location: built on flat area, built on a hillside with downhill access route, and built on a hillside with uphill access route. For simplicity, only the building on the front edge of the plot is shown, and not the two other cases of building location. Viz. in the centre and at the back. The passage from the monocellular to the bicellular overlapped organism (A) occurs on the hillside by gradually adding floors to the building. This leads to the separation of working activities, stables, and storage, from domestic life. The cell on the lower level is partly underground and has only one source of ventilation, in the centre of the entrance wall. Since this is exclusively a ventilation device, it can be said that the window has made its first appearance. Partly to accommodate this opening, the access to the cell on the upper level is moved to the side. The next variation is the duplication in depth along the slope (B), as any other addition would be not poss ble uphill (B3). The organism develops in height with the stacking up of several cells and the introduction of an interior staircase adjacent to the wall opposite the entrance. The lateral duplication of the cell occurs in the earliest examples, usually in proto-urban settlements, leaving an ambitus of around two feet. The subsequent mutation, however, already encompasses the common shared wall. The lateral aggregation of cells presupposes the interjection of the staircase and, eventually, the roof pitch rotates by 90 degrees. In the elementary type, if the roof is not flat, its pitch is usually perpendicular to the contours; in other cases (C2), valleys are eliminated to facilitate drainage. Developments in height (E) are affected by the positioning of the interior staircase.
the building type the elementary type There is a limit beyond which our enthusiasm or our credible seriousness lose their value, and that limit is the concept of the house acquired over time: the house viewed in its definite and particular generality. And our duty then must be not to contradict that view. Heinrich Tessenow Housebuilding and Such Things Such a comparative analysis reveals that even the most complex residential types evolved from the single cell house, that is, from the mono-cellular type, assuming the mono-cellular type can be accepted as the most primitive architectural model for family life. I call this the elementary type, which starts out as an undifferentiated room, and is then articulated in the interior according to function, then developed through successive duplications. Finally it is enlarged by hierarchizing the different spaces to become a new leading type. The mono-cellular type usually consists of a quadrangular room 25 to 35 sq. m. in area; its rectilinear walls give it the intrinsic ability to be replicated, aggregated, and divided up. The ratio of its two sides varies regionally from 1:1, 2:3, and 2:2.5, etc. A single opening simultaneously serves to ventilate, illuminate, and provide access. In the hierarchy of complexity - element, structure, system, organism - the mono-cellular unit occupies the level of systems and operates autonomously, either individually or within another system. In the history of transformative processes, the mono-cellular type represents an achievement of mankind equivalent to the invention of the wheel. In addition to the technologically important act of joining walls at a 90-degree angle, its ability to aggregate this type leaves all circular-plan types derived from the hut, such as the trullo and the nuraghe, behind. It is therefore not surprising that in marginalized areas the quadrangular type does not occur in pre-historic times but not until the Middle Ages, when these areas begin to show an autonomous cultural development. In some cases, however, a quadrangular type never occurred - for example, the trulli of Alberobello in Apulia show an elaborate technology in stone that allowed roofs to be built that sit on a circular plan. A compromise to aggregate several cones in larger units is achieved by surrounding the base of each cone with a fluid connecting space. Analogously, in Yemen we find in the pre-Islamic period a rural type with a tower and circular plan called a nobah. In the cities, the tower type keeps the same vertical distribution of the functions, with the only exception being the stairwell, which moves to the centre, but in order to favour the aggregation in rows or clusters, it transforms the circular plan into a quadrangular one1. In reading more complex residential types, because the mechanism of doubling and specialization hides its tracing, it is sometimes necessary to postulate the mono-cellular unit. Even when it is geometrically legible underneath a complex stratification, the mono-cellular unit may have lost its original function.2 Normally, however, it is recognizable as the
67
68
after amnesia Fig. 28 The trulli of Alberobello, Italy, are the transition from the isolated circular hut to the aggregation of elementary cells with orthogonal walls.
smallest module, that is, a room, but it can most easily be recognized in all serial building generated by the repetition of a spatial module. In these highly serial special buildings - even in such disparate expressions as Gothic and Arab - where articulation is achieved by the scansion of the support system, the span, comparable to the mono-cellular unit, achieves a typological autonomy of its own that at the hierarchy of systems complexity corresponds again to the level of systems, as shown in Choisy’s tables.3 The Arab mosque provides an excellent example. In its simplest version in an isotropic building tissue, it is nothing more than the theoretically infinite repetition of a mono-cellular unit in all four directions of a plane. To demonstrate that this is not just a theoretical distinction but a spatial organism and temporal context in history, we can point to some concrete examples. The countryside, long the subject of study of the French Annales School, is where we can most easily find these elementary matrices, for in the city they are buried under complex stratification. Rural building is an expression of an existential, utilitarian architecture that responds to everyday problems: shelter, warmth, the storage of food, the protection of domesticated animals. It avoids gratuitous innovations, uses the simplest techniques to ensure a certain stability, and efficiently meets the most basic of family needs. The relationship between dwelling and place (topos) was first consolidated by the peasantry, so it is always profitable to turn to the countryside for an initial reading, regardless of scale. The first urban building systems, when they are the product of internal metamorphosis rather than colonialism, are influenced by the rural hinterland.4 The matrix
the building type type (i.e., the leading type of the first urban settlement) is the same as that of the hinterland that has lost some of its rural features like stables and rural annexes. Ethnographic studies and archaeological excavations in the countryside provide many examples of the medieval elementary type common to the Mediterranean region. For example, the leading type in the Alpujarras Mountains, the last Muslim stronghold in Spain in the medieval period, is a windowless, freestanding rectangular monocellular house. W. Giese traces it to the Moroccan Berber house, the taddart.5 Similar, though more developed, structures have been discovered by P. Allart and M. C. Delaigue in the Granada region at Capileira. They consist of a rectangular room roofed by a terrace in launa, locally known as an Andalusian nave.6 Much less common, according to André Bazzana, is the square-plan type found in Andalusia: a 5.8 m x 5.9 m structure has been excavated at Mollet, and two parallel rows of 6 m x 7 m cells have been excavated at Marinet.7 The Berber houses of the Zbal tribe in the northwest part of the Rif in Morocco are arranged beginning with a cell of around 250 x 550 cm, while the Ben Said tribes use a 220 x 480 cm cell. The room of the Kabilis in Algeria comes close to a harmonic relationship of 2/3 (fig. 33). In the mountains of the Aurès in Algeria, the module is a traverse space 200 cm in width, usually doubled with piers and beams. The relationship between the two sides of the cell becomes more noticeable in the ksour of the Atlas Mountains. The fortified houses, or tighremt, characterized by four angular towers, have long, narrow rooms in a 1:4 relationship. A variant of the tighremt lays out the narrow rooms in a T or a cross plan, freeing up large rooms and establishing the premises for a hierarchical plan with served and service spaces (fig. 29).8 F. Ragette also calls our attention to a Lebanese example: the singlespan masonry house at Baalbeck, either square or rectangular in layout.9 The quadratic Palestinian cell covered with a dome is based on a mixed construction system of masonry with cross-vaulting, built on a base of center-supported wood formwork of about four meters in width, subsequently carried out in stone with a slightly parabolic section.10 In the complex of courtyard houses in Qasr al-Tuba of 743 CE the building comprises a single large room covered by a barrel vault. In the Palestinian cell and in the houses of Qasr al-Tuba, in fact, form and structure coincide. The possible main lines of evolution of the mono-cellular type are shown in synthesis in figure 27. The table is useful for understanding the anthropic growth of the type. A double-entry grid explores evolution in relation to access and slope - for instance, whether access is from a valley side or the top of a hill. For the sake of simplicity, analysis is limited to a building constructed on the edge of a lot; the two cases of a building constructed in the middle of a lot have been omitted. The transition from the mono-cellular organism to the superimposed bi-cellular one is based on increased housing needs, and takes place vertically on the slope; it involves the separation of work-related activities stables, storage area, tool shed - from domestic activities. In examples taken from both Kostur-Kukes or Dhernmi-Vlore in Albania (the phenomenon is perfectly legible even in an urban setting)11 and Bosnia, the lowest story is a semi-underground cell with a single opening and ventilation source in
69
Fig. 29 Ait Hamid, Tigremt. Single-family house with 30-meter sides with four corner towers that reaches a height of 24 meters with a cruciform plan and double system of vertical ascents. (From F. Ragette, Traditional Domestic Architecture in the Arab Region, Sharjah, 2003, pag. 112)
70
after amnesia Fig. 30 Two building phases of the Palestinian cell (from Ragette, op.cit. pag. 42) that consist of a unique system of masonry cross vaulting with mud cover constructed over centre-supported wood formwork.
the middle of the only free wall. There is a side door on the ground floor. For the first time there is an opening for the sole purpose of illuminating a room; it is set in the center of the wall that receives the most sunlight. A doubling in the depth along the slope introduces the problem of the distance between the outside door and the stables, and is resolved by a steep external staircase on the short valley side that connects the two cells. The location of the staircase conditions the placement of the openings: the stables can be reached by passing under the first-story staircase landing. In the third case it is impossible for the organism to double in depth towards the hillside so it grows vertically by layering cells and introducing the stairs adjacent to the inside wall opposite the door. The origins of the tower house are established in this way. There are examples from Morocco to Yemen, all the way to Albania where, in the mountainous area of Puke, we find squat three-story towers above ground: the first floor is used for stalls, the top floor, rendered autonomous by an external staircase, is used for the family residence, while the middle floor, called “friends’ house”, is used for guest quarters. In very early examples of proto-urban settlements lateral doubling occurs, leaving a separation ambitus of approximately two to three feet, though a later type already includes a common wall.12 The lateral joining of cells assumes the interjection of a staircase and may involve the rotation of the roof framing. In fact, the roofing of the elementary type, when not flat, tends to locate the ridge perpendicular to the contours; the valley is usually eliminated in order to facilitate runoff. This decision is linked to the organic level of the local culture: if serial, it might cover each individual module separately; if organic it might favour a unitary image for the structure. A further doubling in height goes hand in hand with the complex problem of where to place the interior staircase. This problem is an old one, especially in western Europe, where the row house was the leading type throughout the Middle Ages. The staircase initially was parallel to the supporting wall and lined up directly with the entrance (causing problems for the framing of the ceiling). Later, in early attempts to create multi-family dwellings, it was rotated 90 degrees
Fig. 31 Tower house in the village of Dhermi-Vlore, Albania ( From Monumentet)
71
the building type parallel to the main structure of the ceiling and ultimately ended up as a dog-legged staircase in its own specific location between two load-bearing walls. “Mountains are ideal places for preserving the past,” writes Fernand Braudel,13 and it is to that stronghold of tradition, the Berber mountains of the African Atlas, that we now turn, retracing the footsteps of nineteenthcentury French folklorists.14 The Grand Kabylya in Algeria - a series of ridges that wind down the top of the Djurdjura mountain chain parallel to the Tizi Ouzou valley - is an area cut off from principal traffic routes because of its daunting topography. Until very recently time stood still for the inhabitants, a people devoted to subsistence farming who have long maintained their ancestral way of life, especially in how and what they build. Within the Grand Kabylya the Beni Yenni tribe constitutes seven villages which are today undergoing a process of rapid transformation, thanks to a steady flow of capital from its emigrants to France. Our primary interest is in this rapid change in a previously stable system. Evolutionary changes in the elementary house type that elsewhere took place in pre-industrial times are only now taking place in this area, with one important difference: rather than seeing a series of incremental changes over a period of several generations, we are seeing many changes at the same time. The Kabylya region is therefore like a scientific laboratory in which we can see mutations which otherwise could only be reconstructed a-posteriori. Like all human endeavours, the synchronic variations generated by individual homeowners as they try to solve distributive problems in difficult terrain never cease to amaze us by their ingenuity.15 The Beni Yenni house reproduces the local family structure around a common area.16 As if to emphasize this relationship between space and social nucleus, the word akhan means both a big house and an extended family. It is also used to distinguish between the house of the head of the family, where all family functions take place, and other constructions in the same courtyard, simply considered rooms. The 2:3 rectangular plan is repeated with few variations throughout Grand Kabylya: it presents a double division in width and height that makes three sections clearly distinguishable. The door is found on one of the two long sides - according to P. Bourdieu17 traditionally the eastern-facing side - and provides access to the largest part of the structure, which takes up two-thirds of the total space. The aguens or thakat, the center of domestic life, is adjacent to the stables (known as the addainin). Within, a sort of loft called a tharichth serves as granary and storage area. In the middle of the aguens a shallow hole in the pavement is used for a fireplace. The main entrance also serves to get the animals to the stables, which are at a slightly lower level. A doublepitch roof covers the structure; three wooden beams are fitted in the main walls and held up by three wooden pilasters the height of the thadekuanth. All this is under a single roof. This apparently simple matrix type is the result of a long specialization process that began with an undifferentiated room. Later variants that result from the process are the separation of the thakat from the addainin by a full-height wall and, in more sophisticated versions, a separate external entrance providing the access to the stables.
Fig. 32 Bicellular perpendicular house with slope in the village of Dhermi-Vlore, Albania ( From Monumentet)
Fig. 33 An elementary cell called an axxam, a traditional house in Kabilia, Algeria.
72
after amnesia
Fig.34 The village of Ait Lahsen in Kabilia, Algeria: ( A) original house, (B) repetition of the cell downhill; (C) lateral extension of the axxam.
This demonstrates that even a single system, such as the traditional kabyl house, results from a long specialization process that begins with an undifferentiated room. Secondly, even in a sedentary society inextricably connected to its traditions and apparently immobile, the type is not a static but a dynamic and operative entity. In conclusion, the elementary cell, overlooked by archaeologists and historians because of its simplicity, is fundamental for two reasons. It represents the primary expression of the social needs of a group, or, the physical equivalent of the minimum family unit. It is recognizable as a module both within a residential plan in the form of a room, and in a special, more evolved plan in the form of a bay. On the other hand, in the meaning of the matrix type the elementary cell represents, in addition to the first step of a long processual route, that distinctive sign of a civilization in its territory which, even if covered by many layers, remains indelible.
Fig. 35 Village of Ait Lahsen in Kabilia, Algeria: lateral duplication, but with a reinforced concrete building. Compare to the building types in figs. 27 ff.
Fig. 36 Village of Ait Lahsen in Kabilia, Algeria: modern extension of a neighbourhood correctly aligned perpendicular to the contours and continuous with the earlier houses. The section shows the levels of the old courtyard and of the countryside.
73
the building type the courtyard house Studies on the courtyard house rest on a tenacious equivocation, one that tends to perpetuate the image of a universal and generic type, indifferent to place and unchanging over time. The work of Orientalists and Arabists, celebrating the courtyard as the heart of the agnatic family, the dar as the place of conjunction of the microcosm and macrocosm, as the place that naturally exalts physical and psychological well-being, has confused the issue further. Concentrating on generic symbolic and functional virtues, they have neglected to analyse the structural and typological components of the courtyard house, to follow its phases of change and to evaluate its geographical differences. It is certainly true that the archetype of the courtyard/walled house, as the delimitation of a territory and first act of building, is universal.18 Thus, if there is no sense in establishing primogeniture for something that is congenital to all humanity, it is necessary to realize above all the fact that every cultural field has developed in a completely original way the generic archetype of the shelter-within-a-wall, whether through the choice of an elementary original cell, or through guiding the branches of the typological process differently. Here we will attempt to define the cornerstones of the typological process of the courtyard house in the Mediterranean area with reference to its components and its behaviour in relation to the building lot, leaving aside for the time being the more complex problem of its aggregation in urban fabrics. In the typological process of the courtyard house there was a critical moment when in a precursor an area around a mono-cellular unit was marked off by an enclosing wall. After that the enclosing wall became a reference point, with an aggregation of more cells around a central space. Unlike the side-by-side placement of serial cells, the enclosure simultaneously suggests the final form of the courtyard house and emphasizes its content. Much has been written about the sacred significance of the courtyard house in most cultures. For example, it has been suggested that the courtyard of an Arab house evokes the Garden of Eden.19 Gottfried Semper20 associated the enclosure with a southern Mediterranean agricultural society that must struggle to coax a harvest from grudging soil and protect it from the elements; G. Buti used linguistics to tie it to Indo-European nomadic people. The type is, however, a generic domestic form of residence which independently evolved in various places from the Egyptian-Sumerian civilization to the Mediterranean, Asia Minor, and right up to the Indus Valley.21 The closed and reinforced courtyard house is thus a product of cultural polygenesis dating to the Bronze Age, and it has endured in the Mediterranean basin in the form of the classical atrium and pastas house to be adopted finally by the Islamicized peoples in the Dar al Islam. A number of scholars have observed that while the courtyard house is the leading type in many regions, such as Padania (the Po Valley), the Maghreb, and the rest of the Middle East, it is not uniformly dispersed throughout the Mediterranean. While it is reasonable to assume that the
Fig. 37 Above, an example of “exterior” courtyard house and below, an example of “interior” courtyard house.
74
after amnesia
territories of the Roman Empire were influenced by the domus up to at least the fifth century, no trace of it can be found in parts of central Italy (including Rome), southern Italy (excluding the Naples area), and Provence. Climate certainly is not the underlying reason-for instance, Milan and Aleppo share the type but not a climate. Perhaps the reason for the dispersed presence of the courtyard house lies in civil discontinuity;22 in areas that had been abandoned by the Byzantine Empire, the revival of building was based on a relatively retrogressive type almost consistent with the elementary cell. Within the Islamic world the courtyard type also responded effectively to the essential Muslim requirements of secluding and protecting women. This explains the easy transition from earlier Yemenite models to the courtyard type by the Umayyads when they reached the shores of the Mediterranean.23 Another hindrance to a clearer picture is that despite recent detailed studies of the Arab house, it is hard to establish clear and definitive typological relationships between the different Islamic areas in the Mediterranean. Ambiguities are still in place. Guy Petherbridge offers an overall explanation for the dispersal of the courtyard-house type by distinguishing two varieties: “The interior courtyard house, where the house encloses a courtyard, characteristic of urban areas; and the exterior courtyard house, where the courtyard borders the house, providing a protected area contiguous with the dwelling units but not enclosed by them.”24 André Bazzana finds that this distinction holds for the Iberian Peninsula; he calls the first “block-like” and the second “attached.” According to Bazzana the difference between the two is a result of a difference in economies: the exterior courtyard was used by semi-nomads (fig. 37); the interior courtyard, patterned after the ksar of the Sahara, was originally inhabited by sedentary farmers.25 Such a schematic analysis is doubtful since it is conducted at a too low level of typological specificity. Petherbridge’s contrast between the two models assumes a stereotyped dichotomy between urban mercantile and rural society which does not allow for a plethora of intermediate positions. The contrast is artificial, and is less the result of an Orientalist mentality than the mind set of geographers, who focus on territory, and historians, such as Torrés Balbas, who concentrate on exceptions, such as the opulent and lavishly decorated homes of merchants and city officials. Furthermore, if we look closely at local typological processes in the Maghreb and Andalusia, we not only discover the inappropriateness of the closed and open models of the courtyard house as an explanation, but also discover that the Arab-Islamic city, whose fabric gives the impression of being frozen in time between the thirteenth and eighteenth centuries, is not mono-typological at all. On the contrary, it is based on a wealth of extraordinary variants which in no way jeopardize the courtyard. Similar misconceptions result from the fact that the idea of an elementary courtyard lingers in the cultural memory long after its physical demise, and is replicated and reused in the same area even after a substantial lapse of time. In conclusion, the exterior courtyard house and the interior courtyard house of Petherbridge are actually two moments of the same genealogical tree. If, on the one hand, those who see the courtyard house as the initial
Fig. 38 House in Essaouira, Morocco on the Rue de la Mellah. The subdivided courtyard tends to change into paths leading to the different building units as a result of the type regressing to its elementary cells. Each family occupies one room, under precarious social and hygienic conditions. In other words, the processes of encroachment can also be defined as the passage from a building type to a tissue.
75
the building type universal archetype of walling in, similar at all latitudes, are correct, on the other hand, introducing a more specific typological level, every similarity appears false. A courtyard house in Jilin in China, a courtyard house in Fez in Morocco or a domus from Italica in Spain are profoundly different, in spite of a superficial similarity (fig. 39). The Chinese house distributes the pavilions within the wall so that they are disaggregated and hardly related to the wall itself, and the four empty corners are never integrated into a unitary organism. The demarcation of territory prevails over the built environment, a common characteristic even for the courtyard houses of Latin American cities.26 In the Magrebian house in Fez as well as Algiers the regularity of the central hollow space, which causes the deformations of the lot to be absorbed into the peripheral rooms, is a factor of organicity, reduced, however, by the layout of the rooms, that jam together like domino pieces. It is not clear whether territorial demarcation or the structure of the house prevails. The domus in Italica organizes its cells constantly turning the corner and ordering them with bilateral symmetry, imposed on the tablinum with the altar of the Lares, which traverses the entire building and equally distributes the hierarchy of the weights and measurements of the individual rooms.27 In the Roman domus there is a natural balance between the territorial demarcation (part of a complex system of geometric division of the land) and the built environment. Among the three examples there exist many intermediate levels. In Dar al Islam isolation has favoured the permanence of more formal schemes of direct Roman derivation such as those found on the island of Djerba in Tunisia in the courtyard houses on a standard lot of 48 by 48 feet (16 meters each side with a surface of 260 square meters) that use the basic type of the menzel. The core of the problem is tied to the concept of organism, or better, that specific balance of the organicity/seriality relationship, which is a genetic patrimony of every culture. We have an almost fortuitous system of serial aggregation in the case of Jilin, of an episodic organic aggregation in the case of Fez, and a total organicity in the last case of Italica. In order to deepen our knowledge of the courtyard house it is useful to make a comparison between the courtyard house and the row house.28 There are fundamental typological differences between the courtyard house and the row house. The row house always lies on a road, faces onto it, and is directly accessible from the outside. Its pertinent area is behind, to be covered in subsequent stages as the area is filled in. The pre-eminence of the building site and the high property value attached to its facing front determine the dimensions of the lot, the front of which is equivalent to the size of the elementary cell, usually less than six meters. The courtyard, on the other hand, has a greater sense of the familiar territory, so that the form of the building coincides with the borders of the territory itself. The same strict relationship to the road does not apply to the courtyard type, so any side of the lot can face the street without interfering with the internal organization of the structure. A first building is even conceivable in which the courtyard is some distance from the road. The unbuilt area dominates the built area because the former must mediate between the inside and the outside of the enclosure as well as distribute the organism from within.
A
B
C Fig. 39 Three examples of courtyard houses: they appear identical, but feature architectural solutions with differing degrees of hierarchy from organic to serial: (A) domus in Italica, Spain; (B) house in Fez, Morocco; (C) house in Jilin, China. Figure 37 above represents an example of occasional seriality.
76
after amnesia
Furthermore, the courtyard house type lacks external openings: they appeared only recently in both East and West from the nineteenth century following a lengthy process and exclusively in urban areas. The fact that a room has only one source of light, that is, from the courtyard, limits its depth to six meters or less, depending on the size of the elementary cell used by a society. This depth can be extended by adding a ground-floor portico into the courtyard with a loggia on the upper level. The building bay can be doubled only by making an opening in the wall that faces the street, which then allows the portico to be enclosed, as in the Mohsen house and in others from the colonial period in Shara Jama al-Drug in Tripoli (fig. 98). Another interesting variant is the type called tighremt of the Moroccan Atlas in the version with elongated cells laid out along the perimeter: reducing the interior courtyard to a well of light a meter on each side, it leaves a very wide, single, continuous intermediary portico/ large room used as a living room by the family (fig. 29). In row houses each house has its own bearing wall inside the perimeter of the lot, thus reducing interactions to the minimum during the process of construction. The courtyard house, on the other hand, requires a higher level of participation, since the wall shared with the units around it must accept the beams of its neighbour. In certain cases the common wall is a planned choice, as at Olynthus, but more often it is the result of an “agreed-upon planning” that presupposes social harmony or the existence of a group of relatives. This is a type of solidarity that is valid, unfortunately, even in the case of collapse, where one house can drag down even its neighbour in domino effect; this type is obvious in the Kasbah of Algiers or the medina of Tripoli, where a collapse of a unit reveals facades with stumps of hanging beams, cantilevered floors in precarious balance, niches and wall decorations, interior perspectives that become facades. In the pattern of courtyard houses, in other words, constructive solidarity prevails over the overall form, while in row houses the form contains the settlement. Our general rule is to begin an analysis with the rural version of a house. The limited changes it undergoes allow for a clearer reading of the matrix type and early diachronic phases. Generally there are two types of rural buildings in the same enclosure: the residence and the annexes, which include stables and a shed for tools and farming equipment. Annexes are usually located on the side opposite the house. When the residential part of a structure in an enclosure is located on the side in front of the entrance, they either line up parallel to the entrance or on the perpendicular side. In North African Arab-Islamic urban houses, traces of these annexes coincide with the kitchen or metbah, the pantry area or bayt el-hazim, and the bathroom, and are grouped together on the side opposite the main bayt.29 The houses may also feature a loft area under the roof, but these are of secondary importance to this analysis. Depending on the size of the portico there are one or two intermediate supports, rendering a first modulation of the structure legible. The dimensions of the enclosure are in no way restricted by the built area: such interdependence is a feature of urban planning. In the Mediterranean, one finds it in both the ancient Greek pastas house, whose frontage varies
Fig. 40 House in the casbah of Algiers at 5 impasse des Pyramides. Despite its exiguous dimensions, this example presents all the components of the leading type. The series of paths shows the whole distributive matrix of the building: entrance, patio, stairway, balconies. Rooms have no access to each other either directly or through accessory spaces, but only through a filter of balcony and stairways.
the building type between 9 and 18 meters (27 and 54 feet respectively),30 and the Roman domus, whose frontage measures between 12 and 18 meters (40 and 60 Roman feet respectively).31 The building unit is oriented to take the greatest advantage of direct sun, which in the Mediterranean basin corresponds to a southsouth-western exposure. Because choice of orientation relates more to production needs than to the building itself, this rule is rigid in rural areas but not so strictly adhered to in towns, although it is still prevalent in the majority of the town houses as well. Given the prevalence of a southern orientation, the built part within the enclosure is either parallel or perpendicular to the road. There are then only three possible access variants for the courtyard house. In the first case, when the building is parallel and adjacent to the route, entry is through the building unit, where in order not to limit the distributive possibilities of the building, it is pushed to a far end. In the other two cases, the building is either opposite or perpendicular to the road, and the entry lies in the centre of the free side. Figures 43 and 48 demonstrate situations that produce different diachronic variations, their relative processes, and the most interesting transitions in these processes. These transitions, in turn, are capable of generating parallel processes of synchronic variants which for reasons of
77
Fig. 41 The Roman domus and its parts. The upper sketch shows the origin of the tablinum in the primitive hut that was in the centre of the enclosure.
Fig. 42 The French cadastral plan of Antakya showing a series of eight Hellenistic blocks embedded in the medieval tissue. All the building types belong to series A of the typological process, successively increased.
78
after amnesia
space are not represented in the various diagrams. In the A 1-2-3 series the elementary courtyard is gradually transformed (diacronical variations) as more and more of its area is covered, so that activities that once took place outdoors begin to take place indoors. This is achieved through the addition of rooms on the opposite side of the initial Fig. 43 Typological process of the courtyard house. The A series synthesizes the variants of the single-family house within the limits of the original plot and with minimum specialization: (A1) entrance and path to the opposite covered part; (A2) entrance and path from the opposite covered part; (A3) layout of the covered part along the enclosure’s longer wall. From left to right, (A1-2) the phasing of the construction of the accessory spaces on the side opposite the cell; (A1,3) consolidation of the accessory spaces with the construction of a portico; (A1,4) unification of the distributive system with a covered passage; (A1,5) tendency to form a courtyard around the centre.
Fig. 44 The B1, B2, B3 series describes the taberna process: in B1,2 two shops are created by duplicating two of the elementary cells of the house on the outside of the plot. The taberna develops autonomously until it constitutes the row-house type along the street, thus behaving as an isolated elementary cell: it could then double its height (B1,3) with the introduction of a stairwell; then grow sideways (B2 and B3) to form an embryo of row-house tissue and, where possible, double its front facade by encroaching with a portico onto the public route (B1,2,3,4). The process continues incrementally towards the inside of the original plot (B5), and the “tabernized” rows are changed into apartments (B6).
the building type cell; these new rooms are then consolidated with a portico, after which the two parts are unified by a covered passage, and the process ends by forming an enveloping courtyard around the centre. Because the example in this series is single family, mezzanines and other possible vertical additions typical of different processes are not included, although it is common to reserve at least one cell with an independent external entry for newlywed couples or house guests.32 In this first series, the house has a rather low level of specialization. This series is represented in an exemplary manner in the Ottoman fabric of Antakya (see figure 42). The plot size even in an urban area is large enough to allow division of function on a single level. Further hierarchies occur when an addition is made on the upper level or by acquiring adjacent courtyards to form a larger building unit. This choice will not be an arbitrary one, but determined by cultural, political and economic factors. An important step in the process of making a house more complex and specialized occurs when functions other than residential ones are introduced. This can take place whenever a residential area comes into contact with a strongly commercial outlying area. The B 1-2-3 series describes this development, which we shall refer to as a taberna process (from the Latin taberna meaning shop), a term first used by the Muratori school to refer to the commercialization of residential property. The appeal of commerce leads to the transformation of the front part of the courtyard house into shops, which is achieved simply by creating a new and direct access from the outside to the existing cells of the house. The taberna develops independently to the point that it becomes a row house on the edge of the road, where it behaves like an independent elementary cell: it is doubled in height by introducing a staircase and is elongated on the sides to form the embryo of a row-type tissue, which recalls the evolution of the mono-cellular unit as in the diagram in figure 27. Where possible, the facade is doubled at the expense of the public road by adding a front portico. This process determined the formation of bi-cellular rows that, once codified, became the basis in medieval Europe for all the expansions in extra-moenia villages. The phenomenon occurred at Ostia Antica, Rome’s trade centre, and was identified and described in detail by Alessandro Giannini.33 Further additions occur as public space is occupied, a universal and elementary anthropic behaviour. This phenomenon was widespread in the Middle Ages, especially in those Arab cities where the classical foundations - the forum and agora - had been swallowed up, leaving a bare minimum of road. In the linear suq of Islamic cities a strip of tiny shops provides the background for a residential tissue of patio houses. In the case of a spontaneous first settlement, if the tissue cannot accommodate bicellular shops on the street front, it will allow demolition and substitution inside the house, ultimately resulting in a change in use.34 The ruins of Pompeii, as of many Roman as well as Hellenistic cities, indicate that the classical city was a tapestry of atrium houses and peristyles (that is, courtyard houses) laid out on regular and partial lots corresponding to the series A and B discussed above. But classical urban remains also indicate the presence of another important type, the insula.35 The well-
79
80
after amnesia
preserved ruins at Ostia Antica show three- and four-storey apartment houses, some with rows of shops on the ground floor that are schematically similar to modern-day condominiums. A clear example is found in the so-called Hercules Flats, a row of double-facing insulae with staircases every 10-12 meters inside the building unit. These types of buildings also existed in Rome, and according to Saint Augustine were in some cases even up to seven stories tall. The insula of Ostia Antica and Rome is the codification of a common type, or it is designed all together, fixing a moment of a long evolutionary process of subdivision of the domus that is essentially spontaneous. Even the rab of Cairo can be absorbed into an insula. This process of progressively filling in a courtyard, which we might call the “insula process,” or insulization coincides with the transition into multi-family residences of the earlier domus. The C 1-2-3 series describes this process: it begins at a side of the courtyard with a row of cells that doubles in height served by a balcony. The stairwell is strategically
Fig. 45 The C1, C2, C3 series represents the insula process starting with the occupation of one side of the courtyard (C1,2) with a row of cells soon to be extended vertically (C2,2). To increase the living space, the single cells eliminate the exterior staircase and adopt a collective balcony served by a common stairwell strategically placed in the portico of the original elementary cell; then the row is mirrored on the opposite side (C1,3). This leads to the construction of a second balcony served by the same stairwell (C1,5). Another version embeds the staircase inside every elementary cell of the two rows, and the cells protrude another four feet on both sides (A1,4).
the building type
81
Fig. 46 Two axonometric sections of houses in Meknès, Morocco, show the nodal role of the wust-ad-dar. In the house below, the court maintains its stereometry intact. The plot deformation is completely absorbed by the rooms towards the outside.
located in the portico of the original cell; by doubling the row on the side opposite, a single stairwell can efficiently serve another balcony.36 In another a-topical version, the stairs are provided for every single unit. If the open courtyard is eventually fully utilized, then it is reduced to a linear passageway. However, if the process begins on the side opposite courtyard C1 and C2, the tendency is to maintain a small square enclosure. The two processes do not take place independently of one another. On the contrary, commercialization of residential space (the taberna process) and the apartment-block phenomenon (the insula process) often occurred concurrently. The insula process was triggered by two opposing tendencies. 1. In a period of political or economic crisis, social upheaval causes a decrease in specialization; as the rich move out, their houses are occupied and subdivided by poor families. 2. In a period of rapid economic growth, the city attracts new inhabitants, leading to land pressures and the need for major housing stock, which must be created by occupying the interior of existing houses. We deduce from this that building increase is a common phenomenon that society is able to govern reasonably in normal conditions, but moments of serious social unbalance can generate pathological phenomena like wild encroachment, or speculative phenomena like demolition or substitution by denser apartment houses.37 In the Arab or Ottoman pre-industrial city the insula process occurred only in the limited cases of great metropolises, such as Cairo during the Ottoman period, because it is connected to proletarization, which is much less frequent in a medina of merchants and artisans. The traditional Arab house expands in a different manner without dividing the courtyard, a pattern compatible with a patriarchal family system which is culturally resistant to any trend toward multi-family residence. In Europe the transformation of the domus led directly to the codification of the row-house type, while in pre-colonial Arab Islam this happened only sporadically. Thus in the Maghreb the insula type was
82
after amnesia
Fig. 47 Incremental variants of the courtyard type: the D1 series represents the addition of one whole module, whereas D2, increasing by a half-module only, shows the merging process of two contiguous courtyards. This can happen in five basic ways: (1) Non-hierarchical aggregation of heterogeneous courtyards; (2) simple pairing, obtained by demolishing the wall previously defining the two properties, without further intervention; (3) nonhierarchical aggregation of homogeneous courtyards; (4) unification of two pre-existing organisms; and (5) enclosure of the courtyard on all four sides.
reproduced imprecisely, and it was also always strongly resisted in North Africa whenever the core of the house, the wast-ad-dar, was threatened by population pressure. Even when faced with a population density of 500 inhabitants per hectare, the house responds there by growing in height and subdividing in an effort to preserve the integrity of the courtyard.38 Although the courtyard itself lost its “cosmic” value (as Hasan Fathy called it) long ago in the transition from single-family to multi-family dwelling, it continues to represent in the mind of society an idea of unity that goes beyond its distributive function. The status of the row-house type provides the additional proof for Islamic resistance to the insula development in the Mediterranean. In the Islamic city the row house was generally considered a lower-class house, while in medieval Europe it was the house of the urban middle class. The row house type nevertheless has a local history along the Arab Mediterranean coast as the excavation of small bi-cellular row houses at Fustat attests, and is found in the form of small cells built on top of shop fronts and lying in a protective commercial screen on the border of a residential neighbourhood or in the form of descending types with elementary cells in tissues of encroachment (fig. 86.2). A type cannot be stretched beyond its limits, and the metamorphosis of the courtyard house has its limits as well. In recent years in surveying some densely populated medinas of North African urban centers, such as Casablanca, where more than 1000 residents are squeezed into a single hectare and domestic life is seriously affected by the high density, the subdivision of space is more and more common - one family into one room -and there is a tendency to divide up the courtyard into thin access strips that can be covered when necessary. These are the seeds of future row houses or apartment houses and will bring with them subsequent changes in the Islamic idea of the house and the city and perhaps Islamic society itself.39 The extreme case is the New Mellah of Essaouira created at the end of the nineteenth century, where every courtyard was transformed into a series of labyrinthine routes even on different levels leading to single rooms, similar to a small village. Series D1, in which a whole module doubles, and D2, in which there is an increase of half a module, describe the fusion of two neighbouring
the building type
83
Fig. 48 The E1, E2, E3, E4, series charts the processes connected with the subdivision of the plot. Each new portion, whether of the same dimensions or not, behaves like an autonomous type. Hence it would suffice to refer to the A series with the addition of a corridor leading as efficiently as possible to buildings towards the inside of the plot. If minimal portions are left over from previous subdivisions, synchronic variants will appear as adaptations of diachronic types.
units. The process is subject to any of five basic outcomes. First, an un-hierarchized aggregation of heterogeneous courtyards; second, a simple joining of two plots by eliminating the dividing wall and making no further modifications; third, an un-hierarchized aggregation of homogeneous courtyards; fourth, a union of two pre-existing organisms; and, finally, fifth, a closing of the courtyard on all four sides. This overall process happens differently in cities in the Maghreb, where the expansion of a house is achieved through the serial addition of another autonomous organism. For instance, a very large house may have three courtyards, but they would never be re-hierarchized as they were in the Italian Renaissance case of serial courtyard buildings being turned into a palace. Instead, the resolution of the whole is achieved by making an opening in one of the walls of the acquired unit, which, if it is smaller, becomes a service unit, or by adding an elevated passage (sabat). Finally, we note the hierarchical aggregation of heterogeneous courtyards and lastly the hierarchized aggregation of homogeneous courtyards. The former situation heralds the great Syrian dar, which sometimes coincides with an entire block, like the al-Hawarina and Gazzar houses, both in Aleppo, and
84
after amnesia
Fig. 49 Nizam house in Damascus, Syria. In reality, this building of the 17th century arises as the coupling of two houses, but without the fusion into a larger courtyard. Series D of the development of the house by acquisition of the adjoining lot is rare in the Islamic world, differently from the West, where it leads to demolition and doubling of the courtyard.
the latter foretell the Italian palazzo for those areas where there is a strong permanence of the courtyard type. The E1, E2, E3, E4 series describes processes relating to the subdivision of a lot. Each new portion, if uniform in size, will behave like an autonomous type, taking into consideration the presence of a corridor that must serve the innermost type without being disruptive. In the case of very small portions synchronic variants occur for the adaptation of diachronic types. The purpose of figures 43-48 is to reconstruct the principle phenomena tied to the typological processes of the courtyard house and not to serve as a universal model, and they are therefore subject to adjustment. It should be kept in mind that every building culture behaves differently and privileges its own itineraries in this scheme; this could be a function of historical processes or of uniqueness of location. For example, the elementary cell can be square or rectangular and of varying sizes. The aggregation of the cells is linked to the level of organicity of the cultural area, an area of organic vocation that prefers to order environments from a route and/or visual axis. The corner cell can be part of a larger rectangular cell as in the Maghreb, or inscribed in a unitary compositional discourse as in Pompeii. All this has important consequences for the higher scale level, the building tissue, theme of the next chapter.
the building type
85
reading the courtyard as an organism of systems of structures of elements: fez The Islamic house displays structural complexity in the organisation of the building. The analytical decomposition of the building organism of grid 1.5.9.13 of fig. 9 can end up being easier if applied to a series of examples belonging to diverse areas or to a mixed building culture like Fez in Morocco, or the elastic-wood culture of Mostar in Bosnia or the plasticity of masonry technology like that of Aleppo in Syria. Analytical decomposition has an immediate demonstrative function, but it is abstract and reductive: in following it, the reader must not lose sight of the fact that every element should be evaluated in its context, that is, the structure of elements, the latter in relation to the system of structures of which it is an integral part, and must not separate the system from the entire complex of hierarchical relationships that make up the building organism (organism of systems). The “ bourgeois” houses of Fez of the 17th-18th centuries and their typological variants are the point of arrival of a long experimental route of the typological, technological and functional components that are the result of a profound artisanal know-how. With reference to construction
Fig. 50 The Dar Sqolli at Fez, plan and section through the courtyard ( from J. Revault, L. Golvin, A. Amahan. Palais et demeures de Fès, II, pag 78 and 86)
86
after amnesia
these houses are built of wood and masonry materials utilised in a fashion that simultaneously exploits structural and thermal properties. The proposed reading concentrates only on the main passages in order not to indulge in a punctilious description of the domestic parts. For a detailed description of all domestic parts the reader should consult the exhaustive work of Jacques Revault.40 If for the purposes of the organisational analysis it is necessary to leave out the accessory spaces that are unessential to our discourse like minor partitions of rooms, encroachments that may close the courtyard, rooms acquired outside of the original plot, and hangings (tahal) or suspended passages (sabat, the typical courtyard house at Fez is a two-storey building with a patio closed on four sides with a double porch or external gallery. It comprises modular rooms called bayt or byt of elongated rectangular form and arranged along the edges of the plot. Secondary rooms like the kitchen, toilet facilities and pantry are in the corners, sometimes served by a corridor. The stair is located in the corner opposite the entrance. In more wealthy dwellings the vertical arrangement is doubled: a stair leads to the quarters of the guests, another to the harem and the terrace like Dar Zerouali, a dependence of Dar Zouiten. The arrangement of the bayts at the ground floor is replicated at the upper floor, where the spaces are served through an external balcony or through the rooms in succession. The terrace, usually reserved for women and children, is closed by a high wall along the perimeter of the house, while a short parapet protects people from the void of the patio. In the mutual collaboration of the rooms, the plan of each bayt is deformed into trapezoidal forms in order to absorb the external irregularities of the urban fabric, but it maintains the walls at 90 degrees along the courtyard in order to preserve the geometric regularity of the central courtyard or patio. The footprint of the patio tends to be rectangular or square and contains one or two axes of symmetry, marked by the position of the doors of the bayt facing each other in the centre of the side of the courtyard, as in the Merinide Dar Ben Segrun. Sometimes a small fountain at the center reinforces the virtual intersection of the two axes. The symmetry of the patio does not involve the whole building (differently from the Roman domus or the Renaissance palace): the doors facing the courtyard are in asymmetrical relation to the single bayt. The relation of the house to the urban fabric is accommodated through the adaptation of the peripherical cells. The patio or Wast-ad-dar, on the contrary with its elementary stereometry, its arrangement of the openings controlled by symmetry, is the center of the composition and gives essential light and air to the house. Moreover, the transition from the well of the patio to the sky, the special treatment of the overhanging attic with green painted tiles and the wood decoration give to the patio the value of an autonomous sub-organism with its own rules of arrangement and construction. In the Maghreb the wast-ad-dar is strongly protected by high walls and building units and several spaces that progressively insulate the center from the periphery, and control access to the heart of the house. The
the building type distributive mechanisms, apart from some variations in nomenclature, are basically identical in form: (1) the alley connects the street to the house, and one or more doors open onto it; in the past it was closed with a door (fig. 63). (2) In Tunisia the driba is a narrow room outside the perimeter of the house which acts as an antechamber. It is usually formed from an old segment of the alley that was covered and occupied by the neighbouring house. One door communicates with the alley and another blocks access to the entrance of the house. (3) The skifa in Tunisia is the room in the house that serves as the entry. It features two unaligned passages in the shape of a bayonet, and conceals the wast-ad-dar even when the front door to the house is open. As a result of these devices, a visitor is intercepted by four successive doors before finally reaching the house. He will not be allowed to see the inhabitants from the outside - reinforcing the separation between the harem (lit., holy, prohibited) and the halal (permitted) (fig. 62). The filtering mechanisms are not always all present: only the skifa is an essential feature of the Arab-Muslim house.41 It has been suggested that the absence of filtering devices in Berber settlements is due to the relative freedom that women enjoy in those societies. In Kabylya and Aurès in Algeria, one enters the house directly from the courtyard. Filters are likewise absent in the Jewish mellah of North Africa..42 While it is true that the hierarchy of the sequences of openings and closings corresponds to models of social relations and tribal alliances in the Arab world, a wholesale equation of these elements with Islam would be rash: the dogleg or staggered entrance is found in all Middle Eastern houses from Ur and Harappa to Tell el-Amarna and even at Delos and Athens.43 There is no trace of it at Olinto or Pompeii: in the first case there are double doors but they are aligned, and in the second the strong axiality of the type is reinforced by a central perspective. The system of structures (fig. 9-10) is a collection of structural elements linked in a reciprocal relationship that allows for the delineation of a minimum unitary space, open or closed. At the same time, the hierarchically ordered correlation of systems (systems of systems) comprises the organism. The system can always be disassembled into its parts: foundation, walking surface, higher floors and covering. With reference to the materials used, a system can be homogeneous or heterogeneous. The homogeneous system, realized with the same material in vertical and horizontal structures, has an intrinsically higher level of organicity. In the courtyard house types of Fez a system of closed and heterogeneous structures is the elementary cell, bayt , a tridimensional module of 220 cm of maximum width (equivalent to a kama or a double arm of 165 cm plus a dra or cubit of 55 cm), whose length extends up to 7-8 meters. According to some scholars, the width is determined by the ancient use of palm wood beams, while for this author it is a transliteration of the rural Berber elementary type called taddart.44 Functionally the bayt has the highest flexibility and autonomy: it contains either the master bed or the living room couches, and is separated from the patio by a monumental door. It can be easily transformed into an independent dwelling unit for a nucleus of the extended family. Quite common in Tunis, and less so in Fez, is the articulation of the
87
Fig. 51 Dar bou Helal in Fez, Morocco. To the left, plans of the main floors; to the right, the axonometric section. The house is made of two units: one is a courtyard closed on four sides with a peculiar symmetry - the doors of every bayt are located on the axis of the wustad-dar; the other is a court acquired later and turned into service area (compare with figure 49).
88
after amnesia
principal bayt with very deep niches. The T-shaped liwan called bait-al-qbu wa al mqasir is still a system of structures that combines a normal transverse room with a central iwan, leaving two service units at the corners. In Fez, it is present in the Lahlou house-palace, in Algiers in the house of Moustafa Pasha and in Tunis it is articulated with three iwans in the Husayn house. The T-shaped plan is found in Tripoli in the Bait al-Mal and in the House of the Pasha (former Sharaitic courthouse) in the quarter to the south of Arba’a Arsat. In both cases they are a kind of independent apartment, used by the owner for audiences, conversations and rest. Similar arrangements are found in Syria in the tarma and riwaq model. The gallery on pillars or columns is an open structure that architecturally enriches the houses of the wealthier families like the Abu Helal house or Adeyal house.45 The lower order sits on posts of masonry that define a porch of minimal width that filters air and light to the more internal spaces. A dense decorative pattern of stucco covers the surface, reducing its weight and making it vibrant in the sunlight. The upper balcony has the same function as the patio and porch of arranging the routes, avoiding the passage in succession in the rooms. The gallery rarely runs along the entire perimeter of the patio; more frequently it is a loggia in front of the main entrance, as in the Dar ben Seqrun house. To proceed to a reading and deal with the level of the structure of elements, it is necessary to introduce the notion of the structural node. When two structures intersect, they form a node, which can be defined as the intersection of two continuous points or one single point of a continuum. A linear node is the intersection of two planar structures, such as the line at the intersection of the exterior wall with the ground and that of a floor slab with a bearing wall. Nodes are also openings, understood as the subtraction of material from the bearing and insulating section. Very delicate nodes of an organism are the attachment to the ground, the meeting of the pier and covering and the pier and ceiling, especially if there is a change in material. In order to define a system, therefore, it is not sufficient to describe individual structures, but attention must be paid as well to their nodal relationships. The examination of nodes, of the solutions connected to the intersections, allows one to decipher the language of a cultural area, because nodes, along with measurements, are the signature of the builder. The structural members are aggregates of elements that contribute to the formation of a system. The timber structure is constructed as a frame, i.e., an open discontinuous structure that works with interspersed supports; the masonry structure is closed and continuous and stands as a permanent element. The system of mixed structures in the Fez houses reached a high level of integration that required the alternate presence of the mason and the carpenter on the site.46 The foundations of the houses are in continuous masonry made of bricks and stone laid in mortar. The vertical structures are closing and continuous; three sides of the house are normally shared with the neighbouring houses, the fourth, the façade, has no relevant openings but the main entrance. The unitary treatment of the surface of the facade with
89
the building type a plaster of lime and sand reinforces the sense of continuity of the masonry envelope. Inside, the wall is more permeable to light and communication: the surfaces become an instrument of the definition of the space and relationships of the courtyard. A polychrome ceramic (zellig) covers the interior wall, continuing the geometry of the pavement. The bearing walls of varying section between 50 cm (ground floor) and 35 cm (upper floors) appear on the level of the terrace, dividing it into zones that correspond to the distribution of the rooms below, a choice apparently without functional purposes, which, in accentuating the tectonic importance of the node, weakens the organicity of the entire masonry box. Of all the combinations of elements, the floor slab received particular attention because it was the most visible manifestation of the building tradition in different cultural arenas. Conceptually the floor slab is the projection of the roof on a horizontal plan, therefore in the cultures that use gable roofs the ridge purlin will be transformed into a master beam, while in those that prefer a flat roof, the floor slab does not have a hierarchical system of beams.47 Timber horizontal structures in Fez have different levels of complexity. The simple slab of a bayt is made with joists (ga’iza) ranged with a span of 15-20 cm and encased at both extremes. On this structure sit floor boards of 1 cm thickness on which is spread a bed of mortar. Above lies a stratum of rubble materials (50 cm for the ceiling and 30cm for the roof), on which are fixed the ceramics or the waterproof mantle. This type of structure guarantees the uniform distribution of weights on both main bearing walls, but can span only 220 cm in width. To increase the span of the ceilings, brackets were a common feature. Even greater spans for special bayt are realized with wood coverings with keel domes (gubba). The simple slab is the confirmation of a long custom in the cultural area of Fez with flat roofs. It is markedly different from that of Essaouira, for instance, where the cultural foundation goes back evidently to the roof of the hut with two gables. In this city on the shores of the Atlantic thuja beams with a maximum span of 3.20 meters carry a secondary structure, onto which is placed a layer of flattened branches (tassuit), the floor comprising a layer of dried leaves and a mixture of mortar and earth for the floor foundation. In Fez the complex devices of the wood covering of the top floor (halqa) contribute to establishing the higher level of complexity of the courtyard-sub-organism. There are two possibilities: the roof is held up by a trilithic structure, made of brick quadrangular pilasters and beams of cedar, which forms a portico in front of the bayt, or, the overhanging roof is formed of transverse and longitudinal beams embedded in the wall, held up by brackets that act as span reducers. Galleries built on pilasters require a more complex structure of the slab in order to create a passage of 60-120 cm in width. In this case the joists are laid on a continuous beam called the “mother” beam, that runs along the perimeter of the courtyard. The latter usually sits on two overlapping beams supported by the columns and cantilevered toward the center to reduce the span. The beams are coupled with elements of the same section
Fig. 52 The wooden system of construction of the porch.
90
after amnesia
and assembled in the form of a box. The timber structure is faced with a cedar-wood strip with floral motifs, turi, or cufic lettering, qufi, to improve its appearance.48 In the case of the overhanging halqa, carpentry reaches the highest building levels with many superimposed layers up to a meter in height: in order, qantra beams at 45 degrees fixed into the wall with the support of span-reducing brackets, that lead to diagonal brackets (made, in turn, of two or more joined elements), on which is placed the intersection at 90 degrees of two mother qantra parallel to the facades (often superimposed on another beam). Last, the layer of joists, which bears the wall of the parapet, and the coping of tiles rests on or is fixed on the main beam. The system of tunnel structures, through the iteration of piers, has theoretically unlimited spans, if not those codified by the type, which at Fez are no more than 18 meters as in Dar Bench Kroun, while the second, with beams projecting at 45 degrees, is limited to small courtyards up to 5 meters in span. In spite of the fact that the contrast between the light color of the masonry and the dark of the cedar timber produces tones of great visual effect, the attractive appearance of the traditional courtyard house should
Fig. 53 Fez. Wooden ce lings in residential types.
91
the building type not hide the strict functionality of the structural methods employed. The sequence of masonry-timber-masonry is a defining feature of the house type. Lastly, we look at the elements and note that the masonry envelope is made of various elements (but those favoured are bricks (lajoun), placed at 1.2.3.4. headers and with various types of textures), and it can alternate courses of bricks, stone and wood. The three main textures - horizontal, oblique and herringbone - can appear in alternating sequence in the same wall. The wall is covered on the outside with a thick plaster and on the interior by stucco, tiles or plaster.49 The wooden horizontal surfaces are placed on components that can always be differentiated tectonically from the untangling of the technological nodes. The qantra beam has a maximum length of 4 meters, while its cross section varies depending on the span of the covered room and its position within the overall structure. In order to correct the optic deformation from below of the superimposed beams, the top one must be higher. The connection joists (ga’iza) with elongated rectangular section 20 cm high and a maximum of 220 cm in length determine both the width of the bayt and the overhang of the halqa. The corbels are always in symbiosis with the beams and vary in size from three quarters to a quarter of the latter. The arched openings of the wall are closed by rectangular double doors 3.50 meters high. The anta pivots on cylindrical hinges inserted below in a hole in the pavement and above in a conical or parallelepiped wooden element, rtej, anchored in the wall. mostar The houses in Mostar that belong to the Balkan-Ottoman family originate from a fusion with pre-existing types, whose original module is a single room called the kuca, corresponding to the hearth, a function that does not exist in the Ottoman house, which generally places it outside the courtyard. Over time, the kuca was joined by another room, which keeps an autonomous character with the separation of entrances, called “friends’ house”, both in Albania and in the Balkans. Superimposed on this type of foundation beginning in the 15th century is the Ottoman house with hayat, creating an organism with a hybrid courtyard, where the relationship between built and enclosed is very loose. There is no formal relationship between the enclosure entrance and the entrance to the house, the built and enclosed areas autonomously following their respective rules, the former in relation to the movement of the sun, and the latter in relation to the route. The house is introverted, built within the enclosure that separates it from the street, but it can gain public space through the protruding structure of the cardak, a kind of bow-window typical of all Ottoman architecture.50 In Mostar, in the process of parcelization, the street front has acquired more value with greater interest for the façade and specialization of the corners, where the blockage can occur either with residential building or with shops, depending on the level of commercialization of the street. Evaluating the distribution of the organism with reference to the entire courtyard-house combination, the components of the courtyard
Fig. 53 Fez. Wooden ceilings in residential types.
92
after amnesia
are the kapija, the entrance gate, element of inside-outside separation; the main route of the distribution of the courtyard, which corresponds to the crossing of the enclosure and which is distinguished from the route axis of special building by its absence of intentionality; the araluk, entrance space, a place for exchange and business. It is a vestibule that introduces to the courtyard an actual avlija which, in the largest complexes, where a marked distinction between the selamluk, male area, and the haremluk, female area, is possible, takes the form of an autonomous courtyard; the peripheral routes that hierarchize the space in a limited way around their surroundings; specialized or antinodal angular rooms or elements of peripheral discontinuity at the intersection of antinodal axes, which tend to specialize and take on their own characteristics, are the subsidiary rooms of the house such as kitchen, bathroom and stall. The parts of the house are: 1. The hayat, the entrance to the actual living space. A transitional space between the garden and the rooms, it is articulated on two floors thanks to a stair, which it always contains. It is the main distributive node of the house, but also the gathering place for the family in the summer months. In rural and older buildings, it is an open or partially shielded space with double-height openings that tends to be absorbed inside the house and becomes sofa as an ordering element of several hierarchized spaces. Because of its spatial complexity, the sofa is assimilable to the level of the system of structures. The interiorization process of the hayat occurs during a relatively brief time between the 18th and end of the 19th century, first through the closing of the room, and then with very large windows. In the beginning, the added cells are kiosks used as rooms that take the shape of L- or U-shaped plans. Finally, the hayat takes a central position, becoming the sofa, the heart of the house. 2. The room or oda is the minimum common multiple of the house, and is characterized by a marked monofunctionality, rendered possible
Fig. 55 Mostar. Progressive nodalization of the corner in the mahalla situated in ul. A. Buca as one can see from the 1881 land register and in the current one.
the building type by the absence of furnishings in the interior. Access is through the hayat/ sofa on a diagonal axis that hierarchizes the rooms through its elements: a cupboard or musandera, obtained by dividing up the room with a wooden panel, is placed opposite the windowed wall and is composed of several parts: the entrance to the room, a filter that mediates the access to the private space; the real cupboard with shelves and wash-basin, while the other side has the hearth and two cupboards. The secija covered by rugs is located under the windows and can take up even two or three sides of the room. In the wealthiest houses, a projecting wood structure extends into the cardak. Differently from the bayt in Fez, the oda, because of its complex spatial articulation, can be assimilated to the degree of complexity of the system. The static-constructive system is mixed, made up of the elasticity of timber and the plasticity of masonry structures. In Mostar the masonry part prevails and the walls continue to the upper floors to create special rooms like storerooms. The enclosure wall of the courtyard and the ground floor are always made of stone. All the horizontal structures of the ground floor and upper floor, the bearing parts of the upper floor, the roof covering, the vertical structures of the hayat, the stairway, the casings and the false ceilings are in wood. The house is situated on a foundation structure and ground support of continuous masonry with a thickness so slight that it appears to rest on the ground.51 The foundations and walls are realized in krecnjak rock and lime mortar. Because of its compactness, krecnjak rock is not trimmed, but used in blocks that have just been hewn. The ashlars of the wall structure have differing dimensions and are not placed in regular rows, with the sole exception of the nodes and corners, where larger square blocks are used. The interior and exterior walls are plastered to compensate for surface irregularities. In all the houses in Mostar the bearing walls, including those of the enclosure, have the same thickness. This fact indicates that the weight of the ceilings is absorbed in
93
Fig. 56 Complex of courtyards of Blagaj and exploded axonometric of the house in courtyard n.1. Thesis project of F. Basile, E. Delia, L. Di Bari, L. Lionetti, M. Loisi, A. Margiotta, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari.
94
after amnesia
equal fashion on four sides, and allows for stone bearing walls of relatively reduced thickness up to 50 cm. This lack of hierarchy in the distribution of loads is found as well even in rooms with three bearing walls of stone and the fourth covered with wood piers with or without bracing; in this case, the wooden wall actively contributes to the static organization of the house. This type of structural organization serves a dual purpose: it makes possible the slenderness of a structure that manifests itself in the marked overhangs of the wooden upper floor, and, through an isotropic distribution of the weight and the relative structural independence of each room, it furnishes a guarantee in a seismic area like Bosnia-Herzegovina. The simplicity and roughness of the stone structures contrast with the carpentry of the wooden parts, in particular the cardak and hayat,52 whose elegance is a unifying element of the Ottoman house and a characteristic of the linguistic declension of the Balkans. The roof of the traditional Ottoman-Balkan house is the most coherent example of the structure of elements, of which two main types are known: one with two gables, and the other with one gable and a crowning pavilion. The covering structure is based not on a system of trusses but rather on a system of two rafters and a chain beam. The double-gable roof with tympanum, in Turkish besik, and two slopes covered by roofing tiles called ceramida, is the most common one, present in the most recent houses like that in ulica Alajbegovica 4 and in ulica Stupceva 7. It is also connected to the simplest typologies with two or four oda located in the most regular lots. The pitched roof with kirma pavilion head, covered with stone elements, is connected to more complex typologies like the house in ulica Biscevica 11 or the Kajtazova Kuca in ulica Gase 21. With regard to wood elements, the piers that cross the hayat are obtained from square pine posts, and considering the average inter-floor space of 2.50 meters, they reach 5 meters. The piers are not placed in direct contact with the pavement of the hayat, but are raised on stone pedestals in the form of a trunk-pyramid. The beams can reach 6 meters in length with a rectangular section. These transfer the weight of the ceiling to the piers through wooden corbels, to which they are attached by nails. The whole of the bearing elements of the first-floor catma comprises a structure in wood elements hidden by a layer of plaster. The characteristics discussed up to this point do not contribute to an easy interpretation of the linguistic aspects of the Balkan house: this is introverted and mute toward the outside. All the curtain walls in wood are not left visible, but protected by plaster. Consequently, the built side that is hidden makes it so the structures do not have a direct formal representation. The ground floor is opaque and massive because of the coincidences of bearing and supported parts and because of a complete lack of hierarchy. The wooden structures of the upper floors, instead characterized by lightness and transparency, show structural elements and nodes and exhibit single roles: a bearing part made up of piers, and a supported part formed by the catma plugging/bracing system and glass and wood casing.
Fig. 56a Mostar. House in ul. Stupceva. Distributive system and static system. Thesis project of G. Farano, C.Giorgio, A. Gigante, R. Pisani, M. Romanazzi, P. Spada, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari.
the building type aleppo The houses in Aleppo-which we could define Syro-Ottoman derive from a fusion of pre-existing courtyard types of different derivations, among which is the Roman domus in a modified version already in the foundation stage. The original module of the courtyard house in Aleppo is a rectangular room called the bayt that is 4 meters wide and of varying lengths, corresponding to a multi-functional room, which is grouped in a rather serial way around one, two, three or four sides of a rectangular courtyard of wide dimensions that can vary between 12-15 × 8-10 meters.53 The entrance to the courtyard occurs through a long, narrow corridor placed in a corner position and never axial with respect to the courtyard, whether covered or open. The grouping at the interior of the lot of the built space occurs in relation to sunlight and the street. The first cells built are those facing west. The house is introverted, built within the borders of the lot that are grouped in a serial manner with the other lots of the block, but it can gain public space through the projecting structure of the kisk, a type of bowwindow typical of all Ottoman architecture, or increase within the block by subtracting cells from the adjacent lot. The increase by doubling the lot is not very frequent in the examples we know, and corresponds to residential building that tends toward the palatial type. The filling up of the lot can occur either with residential architecture or with shops, depending on the level of commercialization of the street. In Aleppo, in the process of the parcelization of the lot, the original structure of the courtyard was often markedly subdivided into several courtyards which, in cases of the halving of the lot, keep the regularity of the plan, such as occurred, for example, in the Bancusa-Jib Qaraman quarter, to the housing units at the cadastral numbers 2194 and 2196, derived from dividing a single lot. In the case of larger parcelizations into three or four housing units, the regularity of the courtyard is lost, and a series of filtering structures form between the courtyard and the street that are increasingly more complex as the new lots formed are further inside the lot and thus further from the street. Representative examples of this phenomenon are, in the Bancusa-Jib Qaraman quarter, the parcelization of a lot into five subunits, corresponding to the cadastral numbers 2200, 2201, 2202, 2203, 2204 or the parcelization of the lot into four sub-units, as in the cadastral numbers 2217 and 2218, or into three sub-units, as in the cadastral numbers 2191, 2192 and 2193. The distribution of the organism involves the entire courtyard-house combination beginning with a long, narrow vestibule that connects the entrance door, element of inside/outside separation, and the real courtyard, and it is represented by the distribution and crossing routes of the courtyard, by the routes that lead to the specialized structures of the iwan or the qa’a, and by the routes that lead to systems of vertical ascent. The vestibule leads into the courtyard which, in the larger complexes where a clear distinction between male and female spaces is possible, takes the form of several autonomous courtyards. An example of the specialization of the courtyard is legible in the
95 Fig. 57 (right page) Aleppo. Table of the progressive occupations of the ground in courtyards of variable dimensions.(All 4 drawings of Aleppo are from the Thesis project of G. Ambriola, A. Dimasi, P. Lestingi, V. Lorenzini, G. Minervini, G. Roppo. Adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari).
96
after amnesia
structure of the house near Bab Qinnasrin, at cadastral numbers 835 and 836, where three courtyards are present: the adamlik or servants’ courtyard, the madafa/salamlik, or guest courtyard, and the haramlik, the courtyard for the family. The peripheral routes of the courtyard, which hierarchize the space to a limited degree in the area in which they are found, lead from the entrance toward the rooms, the bayt, placed on a raised level with respect to the courtyard. The corner antinodal rooms, or the elements of peripheral discontinuity placed at the intersections of antinodal axes, which tend toward specialization and assume their own characteristics, are the first to be ceded in the process of parcelization of the lot, since they do not face directly onto the courtyard. These are the accessory rooms of the house, like the kitchen, bathroom, and entry, or iwan. The spaces serving the iwan always have direct access from the iwan itself. The iwan is a structure of the house, a loggia open on one side and oriented toward the direction of the cool north winds, and serves as a receiving and gathering area for the family in the summer months. This is an environment raised about 50-70 cm. with respect to the level of the courtyard, and it is covered either by a wood floor decorated with wood beams that are circular in section (tiwan), in the same manner as several of the internal environments of the house, or, in later examples, by cross or barrel vaults. The two servant areas called qubba are joined to that loggia, having their pavement at the same level as the other bayt, which are used as living spaces or bedrooms,
Fig. 58 Building deta l of the stone dressing.
the building type and which together form in plan a structure similar to a simplified T. The position of the iwan with respect to the court often determines an axis of symmetry, indicated by the presence of flower beds or decorative displays in the pavement. The iwan’s specialization can be read in the façade through a wall parameter hierarchically prominent with respect to the height of the walls of the qubba or the bayt and which opens into a pointed arch delimiting the space of the iwan. Symmetrical windows at the dividing axis of the iwan at its sides signal the presence of the two servant spaces. The specialization of the iwan gives life to the riwaq, an open loggia often facing west and placed on the upper level of the house and which has the same aggregative structure as the iwan, as for example in the house of the Bayada quarter at cadastral number 663. In more specialized buildings, the aggregation of three iwans around a servant space gives place to the qa’a, a winter reception room located generally along the northern or eastern side of the courtyard, but lacking a fixed orientation. For its spatial complexity, the qa’a is assimilable into the level of the stem of homogeneous structures. The central space with cupola, resting on a fenestrated drum, is called the ataba, placed on the same level as the courtyard and passageway between the interior and exterior of the system, and serving the three iwans gathered upon it. The ataba is a non-specialized polar system featuring a fountain corresponding to its vertical axis. The planimetric conformation of the qa’a is given by the T-shaped aggregation of the three iwans around the ataba and by two servant rooms between the iwans (fig. 59). The process of specialization of the qa’a occurred in a period between the 17th and end of the 18th century in correspondence with urban nodalities, and in particular in the Bab Qinnasrin neighbourhood and in the Christian Jdeide quarter. Examples of houses with the qa’a can be discerned in the housing units at cadastral numbers 2540 and 2528 in the Jdeide quarter or in cadastral numbers 793, 835-836, and 837 in the Bab Qinnasrin neighbourhood. The room or bayt is the minimum common multiple of the house and is characterized by a marked pluri-functionality of use, rendered possible by the presence of low couches placed along the sides, used as furniture for a living room or for rest. The access to the bayt is from the courtyard, with respect to which it is raised about a meter. The passage between the inside of the bayt and the courtyard is through the ataba, a square or rectangular filter that is lower by about 30 cm with respect to the walking surface of the bayt and can be of varied dimensions. The bayt, covered by wooden ceilings, has a symmetrical masonry division along the two long sides, parallel to the courtyard or at the confines of the lot. This division is seen in the courtyard in two orders of superimposed windows in an order of main openings, having the function of aero-illumination, and a secondary order that has the sole function of illumination. A series of wall niches, used as screened cupboards with wood panels, part of the system of boiserie covering the walls of the room, symmetrically marks the axiality of the windows opening onto the courtyard. The Aleppo house can also have a second bayt level diachronically
97
Fig. 60 House in the Bayyada quarter, later on transformed into a madrasa, land registry n. 554.
98
after amnesia
99
the building type
Fig. 59 House in the Jdeide quarter, land registry n. 2540.
100
after amnesia
the building type
101
Fig. 61 Corner house in Antakya. Specialized type with doubling of the body of the building. The formal autonomy of the upper floors of elasticwooden structure, even rotated with respect to the lower wall container.
grouped and rarely along all sides of the lot, the entrance to which is from galleries and systems of ascent within the courtyard, shaped by overhanging stone steps. The construction of bayts on the upper level occurs over time according to a progressive aggregation that often begins with the construction of an upper floor along the south or east side of the lot. The diachronic construction of the upper floors is legible in several ramps that give access to the various bayts. The specialization of the bayt leads to the mourabba, a space on the upper floor of the house, along one or two sides of the courtyard, facing east and richly decorated, which is used as a reception room. At the lowest level, entered by short ramps that begin in the courtyard, the bayts have their correspondence in the qabu, semi-interred rooms used for storage or as living rooms during the summer. These spaces, covered by cross or barrel vaults, open onto the façade almost at the level of the courtyard floor through a system of axial aero-illumination with respect to the window openings of the upper floors and define a third set of openings on the façade. In Aleppo the house very often has a level lower than the semiinterred one, the true quarry or mgara from which the stone material for the construction of the rooms is extracted, as for example in Beit Wakil in the Jdeide quarter. The static-constructive system is mixed, made up of prevalently plastic-stone but also elastic-wood structures: in Aleppo, in fact, the masonry part of the surfaces and stone decorations is prevalent, and only the horizontal elements are realized in wooden structures. The
Fig. 62 The sequence of the mechanisms of separation in the Maghrebi traditional house: soqaq, driba, skifa.
102
after amnesia
same is true for the cupboards and framings, or kisks. The house is based on a foundation structure and ground support in continuous masonry, comprising the walls of the qabu, which has a considerable thickness. The foundation and walls are made of local calcareous stone extracted from the quarries below the semi-interred rooms. The ashlars that form the walls of the semi-interred floors are less regular than those of the exterior wall parameter of the façades on the courtyard. In fact, the walls of the bayt that face the courtyard have a double parameter, and the external one, which forms the group of doors and windows, is very regular. The interior parameter is less regular and is plastered, and constitutes the actual wall thickness. In the houses of Aleppo the walls that define the lot have the same thickness as the walls that make up the courtyard. Along the edges of the lots we can read a double wall thickness, the result of the longitudinal aggregation of two contiguous walls, each of which belongs to one of the two adjacent lots. The weight of the ceilings is absorbed equally by the wall facing the courtyard and the wall along the enclosure, and this situation results in the ability to have stone bearing walls with thicknesses up to 80-100 cm. This type of structural organization through the relative structural independence of each room, and the system of double walls acts as a guarantee in a seismic area like the Aleppo region. The refinement of the wall parameters facing the courtyard has a correspondence in the rich decorative and formal system of the openings onto the courtyard, in the decoration of water spouts or in the decoration of the wall parameter itself, called the ablak, comprising the alternation of white and black stone ashlars, or, finally, in the decoration of the wood beams (liwan) of the ceilings, extracted from materials from the Barada basin. The liwans reach 4 meters in length and empty out directly onto the walls. The roof of the traditional Ottoman-Aleppian house is flat: a wooden surface insulated with earth and covered by a stone surface. The Aleppo house is introverted and mute toward the outside. The only linguistic adjectives are legible in the decorative display of the entrance portals, figuratively emergent with respect to a non-plastered wall parameter and featuring ashlars that are regular and trimmed. Instead, the wall structures facing the courtyard are characterized by regular square ashlars and rich decorative displays that make the bearing parts of the wall structure legible. The symmetry and seriality of the living cells at the various levels also determine the symmetry and seriality of the composition of openings along the facades of the central courtyard. In synthesis, Fez expresses a mixed cultural construction area that did not develop sophisticated systems in the wall area after the 14th century, but only in autonomous wooden displays, so that at times the two systems work in an incongruous manner. Mostar, too, is in a culturally mixed area, but the materials are rigidly separated vertically with a solid basement in bearing stone and a light wood structure that is carried. Aleppo is in a stone area that has kept up to the present day a very refined building tradition both for piers and vaults. From the distributive point of view, the Fez type centers on the patio, the fulcrum of the house onto which all the rooms open and over which the galleries of the upper rooms look out. It always keeps its regular geometrical form, while the deformations
103
the building type of the fabric are absorbed by the individual rooms. On the other hand, in the Mostar type the built part is autonomous and decentered with respect to the courtyard, which contains only service functions. What is more, the former is organized by a very hierarchized nodal room, to which all the routes and collective activities of the family refer. The Aleppian type has a more organic relationship with the courtyard, even if it does not reach the formal autonomy of the Fez type. From the structural point of view, the Fez type separates the structures: the stone base with the rooms arranged in pinwheel fashion works with a system of boxes with parts of walls in common and loads not equally distributed, while the wood cage is hung on the interior facades of the wast ad-dar or rests on piers. At Mostar the masonry structure functions with boxes with loads equally distributed to the four walls, while the upper wooden structure is braced. In terms of legibility, the Fez type is closed at the outside from the contiguous houses, and is strongly unitary inside because of the presence of one or more axes of symmetry, and very expressive because of the decoration of the wood structures. The Mostar type presents a partial external elevation and an internal one, the main one on the courtyard. The plan is tripartite and strongly hierarchized by the hayat and the possibility of a central axis of symmetry. If the external openings of the Aleppo type are rare, the rhythm and hierarchy of the openings onto the courtyard with cut stone decoration are specific components of the local language. Axes of symmetry are rather rare in ordinary houses as opposed to the large homes of the merchants of the 18th century.
Type 1: Access to the courtyard through a corridor
Type 2: Access to the courtyard through another room
Type 3: Access to the courtyard through a path
Type 4: Access to the courtyard through a transition room (skifa)
Type 5: Direct access to the courtyard
Fig. 63 Essaouira. Separation mechanisms between the courtyard and the public space in the block between derb el Majboud and Derb Chabanat (Thesis project of M. Barbone, A. Brancaccio, S. Mastronicola, M. Minerva, N. Moschetta, M. Stefania, adviser Att lio Petruccioli, School of Architecture, Polytechnic of Bari)
104
after amnesia
mediterranean typological processes In order to navigate the archipelago of typological processes of the Mediterranean house and give order to the complex relations bestowed by history, it is useful to clarify the concept of cultural area. A house type is a univocal expression of a culture, anchored to a territory, in which a social group has built its own identity. To say that the residential type does not travel is a partial truth that can not hide the osmoses and interactions among cultures. The experience of linguistic disciplines demonstrates that relationships and outpourings are more intense in contiguous areas; and, on the contrary, the more impenetrable borders are, the harder contacts and transitions are, especially in those areas called refuges, cut off for centuries from the great territorial routes, like the Basque country. The relationships are never horizontal, but hierarchized: in one region there will be leading sub-areas that are economically and socially (today, an administrative centre of a plain) and sub-areas that are more backward, more closed, like the Alpine valleys. Thus it happens that the local “bourgeois” class in forming the idea of the common type will in part refer to the residential types of the administrative centre. The concept is relative; all one has to do is widen the scale of the region to include a vaster area and the administrative centers become, in turn, culturally indebted to a higher leading area, the capital of a state. If, in the 18th century, the plateau of Medea, the piedmont of Blida and the coast of Tipasa in Algeria were dependent on Algiers, in turn Algiers, like Damascus, Tripoli and Baghdad, regional capitals, were very mindful of the fashions of Istanbul, capital of the great Ottoman Empire. My first reasoning about the typological processes of the Mediterranean house would seem to confirm a division between the East and West of the Mare Nostrum, which the Roman Empire had cancelled out for seven hundred years. In southern Spain and in the Maghreb the patio house prevails, which is well represented by the house of Fez, which we analysed in detail. The unifying factor, which favours recognisability, is the taddart, adopted as an elementary cell, which has the power to survive in geographical and temporal variations. For the same reason, the abandonment of the 5 x 5m Italic square cell at the time of the Visigoths has rendered difficult the reconstruction of the continuity with the Roman world. Of all the regions, the Maghreb is the one in which domestic architecture is less known. Recent ethno-archaeological studies like those of E. Laoust on the inhabited areas of central Morocco54 have allowed for the isolation of the taddart, the long and narrow cell, that varies little from one tribe to another, and which in the Magheb enters as a component structure, bayt - equal to the T-shaped room of Eastern importation - in every urban and rural house. The taddart, home of the transhumant, is nothing more than the transposition in stone of the nomad tent, laid out along the perimeter of the courtyard, like the tents of nomads around the douar, leaving space for animals in the centre. On the other hand, even for large fragments it is possible to join some regional processes like Syria and Lebanon to Ottoman Anatolia and to the Balkans up to Venice, according to a thread
the building type based on research of centrality and tripartition of the rooms of the house. Houses in Venice, Lebanon, Turkey, and Cairo will be used to trace typological processes because their evolution can be reconstructed reasonably well from the beginning. The first three cases present similarities that are not casual, but dealing with them would be a book in itself. Aside from its historic relationship with the eastern Mediterranean, Venice is a key area because it offers a clear reading of typologies and urban morphology. The work of Saverio Muratori, Paolo Maretto, and Gianfranco Caniggia55 brings an unparalleled wealth of information to the subject, from a systematic survey of the city’s tissues and buildings to a refined theoretical analysis of the data. A peculiarity germane to the layout of a Venetian house is that it tended to remain unchanged over time due to the special building techniques required by its unique foundation. The high cost of a foundation on wooden piles discouraged variations that involved moving the supporting walls. This and the natural resistance of property lines are the reasons why the so-called Byzantine crossing-hall layout has remained almost intact in spite of stylistic trends over time.56 The first building type, the elementary domus, is identified in a room set against the northern wall of the enclosure and joined to a distributive element, the portico (portego). Planimetric location favoured this solution over two others, for these were houses belonging to fishermen, and the courtyards served as a transit point between water along one side of the plot and street along the other. The concern to connect the two sides of the plot determined the character and the peculiar typological problem of the Venetian house.57 The type was at its most developed when a second story and a squero (a service space used to store and repair boats) were added on, and the front of the building was closed up. At this point, the portego lost its original function and was transformed into the sala veneta or crossing hall, open at either end: it then served a distributive function rather than providing light. The placement of street access was symptomatic and can still be read today in either the zone of the portego or the courtyard, which was later covered by successive building bays that were often open loggias. The original bay, however, always remained solid and can be easily recognized even from the outside elevation. This would explain the recurring asymmetry of many Venetian facades such as those in Campo Nazario Sauro, Lista di Spagna, and the San Stae area. In other cases, most notably the Ca’ d’Oro, vigorous attempts have been made to restore a pseudo-symmetry.
105
Fig. 64 Typological process of the Venetian house. (A) original leading type with the long side of the building perpendicular to the waterfront; (B) addition of the portico; (C) addition of one floor; (D) addition of the squero; (E) encroachment of the front; (F) recomposition of the L-shape courtyard house; (R) double-body house (called fondaco); (S) transformation into insulas with small courtyards. (Redrawn and modified from Caniggia, Gianfranco, “La casa e la città dei primi secoli”, in Paolo Maretto, La casa veneziana nella storia della città [Venice, 1986], pp. 28-29)
106
after amnesia
An early development in the insula process established the mercantile type, a combination house-warehouse with a front loggia; its systematic occurrence led to the important development of the street-courtyard (callecorte) type. When considered an independent type, its use is evident in all of the city’s major planned projects: the Rialto, the Cannaregio waterfront area, and the working-class housing financed by the Republic in the 15th and 16th centuries.58 This type is an extremely functional combination consisting of a street segment running from the water to the main overland route flanked on both sides by row houses. Once codified, it was used in all Venetian-influenced Mediterranean settlements, including Dubrovnik. The similarities between this tripartite layout with a central crossing hall and both the Venetian plan and the Byzantine triklinium are striking. F. Ragette conducted important research on the Lebanese house59 in the 1960s that allows us to reconstruct a typological process in that area. This work is important for its breadth - it covers numerous midnineteenth-century examples, the layouts of which betray a very archaic substratum. Unfortunately the collected samples range over too large a geographical area. In Lebanon an elementary type with a vaulted square or elongated plan can be identified in many rural examples. Changes in the type are generated either by doubling width or height. The first case can lead to a serial aggregation of up to five cells, with the addition of an optional front gallery. On its own the gallery, with origins in the Byzantine tradition of the doxatos of Constantinople and various wood eliakos,60 is not a building type; instead it is cantilevered from
Fig. 65 (Left) Anis Haddad house in Mtain, Lebanon. Axonometric section. Typical example of a central hall on the second floor, revealed by the loggia on the facade. (Right) Sheik Adil HusnEddin house in Moukhtara, Lebanon. Plan of the ground floor and axonometric section.
the building type
107
Fig. 66 Lebanon. From the elementary cell to the house with cross-shaped central hall. (1) Abd el-Baki house in Ain Baal: double elementary cell and replica with loggia on the upper floor. The loggia itself is not a type but a synchronic variation. (2) House in Ain Touffaha: split in two by the cells and formation of a central square hall. (3) Mansour Daniel house in Brissa: doubling in depth and formation of the crossing hall with windows on the two opposite facades serves all the other rooms. (4). Muallim house in Kfar Hazir: secondary access creates two systems of distribution, the crossing hall is specialized with formation of an alcove. (5) Adel Tay house in Kfar Him: next passage with a T-form room. (6) Hanna Shikhani house in Bikfaya: mature implant with central crossing hall, part of which serves as a vestibule.
the building, and at most is a system with autonomous typological merits and low specificity, comparable to the structural bay. The crucial point is where, even in rural houses, the intermediate cell became specialized and gave rise to the iwan, locally know as liwan, a room closed on three sides, and located either on the main level or an upper story. It is an architectural element known throughout the Middle East and is possibly of Persian origin.61 The iwan played a central role in the formation of the Lebanese house. Because access to the house did not depend on it, once an external or side staircase fulfilled that function the iwan became the ordering element for the whole house. Its success was most likely due to the role played by its central space (a constant even in the variants), the plays on perspective, and its great flexibility - the result of functional indifference that also limited its growth. There are two basic categories of liwan and cell combination: The first involves the association of elementary groups around the liwan. As the point of departure for the geometrical axis, the liwan establishes hierarchy and scale: some variants have one, two, or three liwans, or two liwans and a portico. This category includes the Mamluk house in Aleppo studied by J. C. David, in which the iwan reaches monumental proportions.62 The second type involves the extension and modification of the iwan into a crossing hall. The iwan protrudes from the facade supported by a series of cantilevered and embedded arches. The room can be reached from either the front through a gallery or from the side by a corridor. This
108
after amnesia
pattern of access results in two important variants - T- or cross plan - both of which were common in the late nineteenth century. The front access creates a main room and liwan separated by a series of arches, upto twenty meters long in large residences. Side access allows the two free sides to accommodate the liwan and a loggia or bow window. The similarity with the Ottoman sofa house is noteworthy, but it is not the goal here to discover a comprehensive system of relations,63 or to answer the age-old question of the origins of the Turkish house. However, by referring to Sadat Eldem’s research in 1930-1945,64 we can reach a limited conclusion with regard to the Ottoman Turkish house of what he calls the sofa type. The Turkish house differs from other Middle Eastern houses in one important and determining way: the elementary cell or oda is consistently the ordering element and is formally and functionally autonomous. It establishes an extremely articulated planimetry in buildings which breaks up the envelope of the house by playing freely with the spaces. The sofa, a product of the predetermined disposition of the oda, is always expressed by a strong topological or geometric centrality and reinforced by symmetrical references. Given this, Eldem identifies the essence of the sofa type in its multifunctional mechanism and introduces an evolving classification. He begins with a hayat matrix type, and then follows with an exterior sofa type with the idea that an iwan can be placed between two odas to form a T-shaped sofa. Next a kosk (kiosk) is built opposite the sofa, with the latter becoming interiorized over time, and in the final phase there is both an interior and a central sofa. Eldem’s methodology, however, is flawed. Though his classificatory scheme is based on an extensive survey of samples but limited to the Marmara Sea region, he compares only layouts of residential main floors and assumes that lower stories are merely subsidiary. He also considers the main house out of context, ignoring gardens and annexes, i.e., the pertinent area, including the muftak kitchen, which is usually separate. He ignores the shape of the lot and the placement of the house with respect to the street. Basically there is a methodological problem in equating inferior hierarchical grades (a story of a building or, better, a sub-organism) with a building type, which should include the construction plus the pertinent area. Is it right to reduce the Turkish house to the sofa? Recently, Deniz Orhun65 in a reading of the Anatolian Ottoman house carried out on the basis of the level of integration of the three principal components and with reference to a low level of typicality, kitchen, hayat or sofa, oda, concludes that the Anatolian houses have a high level of integration of living space, while the segregated spaces would appear to be the oda and the master bedroom (basoda). The distribution of routes occurs according to a treelike pattern with spaces tied to one another with no possibility of alternate routes. To this hypothesis we can compare the examples of the Sea of Marmara. The houses of Urfa, Kayseri, Erzurum and Mardin, on the other hand, have a high level of integration in the kitchen with a distributive plan of spaces connected to one another by more than one connection. A separate case would be the houses of the eastern Black Sea, a transition area where the type with salon would belong to the first case and that with
Fig. 67 Typological process of the Ottoman sofa house as documented by Sedad Eldem. Typological series with real examples from the type with external hayat, to the type with iwan and rectangular hall up to the phase of the introduction of the hayat at the centre of the house. (From Sedad Hakki Eldem, Turkish Houses Ottoman Period [Istanbul, 1984], vol. 1, p. 34)
the building type hayat to the second. The first type diffused in Anatolia would have been exported by the Ottomans in the Balkans. We can add that after 1453 in Istanbul this generic type lost its character because of successive specializations of the external hayat to the internal hayat, and then symmetrical up to the sofa, as respresented by Saddat Eldem. Having finally become the common type of a bearing area, with the Tanzimat it spread over the entire eastern Ottoman Empire from Albania to the Middle East in the 19th century, favoured by the very mobile and specialized Armenian, Italian and Maltese workmen. We can not argue for its being a direct descendent of the Roman atrium house like all the other types analysed (a thesis tied more to an ideological position than an objective examination), because we can not indulge in the hypothesis to reinforce the findings for a shared fundamental type owing to the great temporal distance between them. The similarities between the tripartite layout with a central crossing hall of the mature Lebanese house, of the Venetian portego, and the Turkish sofa are striking. Excluding important cultural migrations in one direction or the other, it is more credible to say that the three parallel typological processes have been elaborated autonomously.66 Cairo provides another important case study, for the complexity of its urban structures results from the long stratification of the tissues and the specialization of the types. With the exception of some aristocratic houses and grand residential complexes, the fabric of the city has been totally replaced. Although a reconstruction of the processes will inevitably be incomplete and in need of adjustment, it can reasonably be traced back, even through the relatively obscure period between the ninth and fourteenth centuries, by using the archaeological and documentary evidence of the waqf.. The relatively recent excavations at Fustat67 led by Scanlon reveal a wide array of types that incorporate earlier discoveries made by Gabriel.68 Gabriel offers the following description of ninth-century courtyard
109
Fig. 68 Fustat. House no. VI. (From K.A.C. Creswell, The Muslim Architecture of Egypt [New York, 1978), vol. 1, p. 126)
Fig. 69 (Left) Qa’a Dardir in Cairo, XII century; (Right) Palace of the Amir Taz dated 1352 (from S. Noweir and Ph. Panerai, Cahiers de la recherche architecturale, 20/21, 1987, pag.30).
110
after amnesia
the building type houses that appear to be based on a model from Samarra in Iraq: Before one lies a composition on two perpendicular axes radiating from a central courtyard, of differing sizes and proportions. A portico with three unequal recesses lines one side of the courtyard: the central recess is wider than the other two and separated from them by two brick pilasters. From behind the central recess under the portico opens a main room, deeper than it is wide and flanked by two rooms which do not communicate with it. On the other three sides of the courtyard and on the axes lie several iwans of varying depths; at times they constitute real rooms but in general are simply modest recesses or even flat niches. The triple-bay portico that leads to the main living quarters is usually oriented east; it never faces due south or north.69 There is a five-hundred-year lapse before similar reliable examples are identified. During this time the reconstructive process is speculative, given the paucity of available fragments, but we know that a new specialized structure known as a qa’a first appears in the fourteenth-century Yasbak palace. With few exceptions it proves to be the protagonist of every variant of the house from that time to modern times. Service areas and annexes are organized around this multi-functional room. The whole forms an autonomous apartment with a fixed central location. At either end of the durqa’a are two iwans; the square room is lit by an octagonal lantern hanging from the vertical axis of a fountain. The floor is lower than that in the two side rooms used as sitting rooms. The walls are rarely homogeneous: space has a tendency to dilate through the deft breaking up of the walls into compartments, niches, doors, windows, benches, and finally mashrabiyyas. In some instances the iwan is articulated in smaller iwans, generating yet another spatial multiplication by creating new central points and inverting hierarchies. A good example of this is the qa’a of the Dar Yasbak, with its two T-shaped rooms joined to the durq’a. The unity of each organism is characterized by the interplay of walls and recesses, placement of the mashrabiyyas, location of the qa’a within the house, and the effect it has on lighting and exterior views. Little evidence exists to help us shed light on the transition from the courtyard houses of Fustat to the Mamluk qa’a: no archaeological remains of Fatimid-era houses exist, with the exception of the schematic plan Pauty published of the four-iwan courtyard of the Sayyidat al-Mulk palace. Although the proportions are grander, the same elements are present: a composition on two axes; portico and salon on one side; open iwan on the other three. I fell that the kuba at Palermo, although dated 1180 might be the missing ring. A more interesting find is a relief on the reception hall of the Ayyubid palace (dated ca. 1240) on Rawda Island. Though the building no longer exists, an accurate description is found in the Description de l’Egypte. It consisted of two T-shaped rooms joined to a courtyard, the two lateral iwans of which were reduced to oversized niches. The dimensions of the center of the room were marked by four sets of 6 x 7 m columns. These dimensions are relatively close to those of Fustat courtyards, and correspond to the durqa’a of fourteenth-century Mamluk houses. The qa’a of the palace at Dardir described by Creswell70 as a closed rectangular
111
Fig. 70 (Right page) Summer hall (qa’a) of a house in the Hauch-Kadan neighborhood in Cairo (from P. Coste)..
Fig. 71 Seventeenth-century Manzil Gamal al-Din alDahabi near Bab Zuwayla in Cairo, Egypt. Note the section of the mandara (reception hall), the position of which intelligently exploits the deformation of the plot.
112
after amnesia
hall illuminated from above by a cupola on the central axis could be interpreted as a later phase in which the courtyard was covered. The hypothesis is plausible and analogous to similar processes: for example, the derivation of the 3-6-3 m. religious structure from the covering of the atrium of the domus, once thought to be an imitation of the model for the 6-12-6 proto-Christian basilica, the first to be codified.71 In the reconstruction of the maq’ad, we do not consider the northfacing upper-story loggia as an independent type because it is incapable of formal autonomy from the Mamluk dar with which it combines to make up a system. A combination of two qa’as and a maq’ad can be found on the uppermost story of the Manzil Zaynab Hatun, an affluent dar from the reign of Qaytbay (1468-96).72 The structure has two entrances: they lead to two separate apartments of unequal size which face the same courtyard: access to the northern one is from the courtyard, to the other - made up of a maq’ad and a qa’a or riwaq - from both the courtyard and the outside. If our reading of the plan is correct, this example represents an important stage in the rise of the bourgeois house. The annexes were intended as rental property.73 As the durqa’a becomes more important to family life, the courtyard’s importance decreases. Eventually it becomes little more than a passageway, along which various activities took place and different forms were imposed. During the Ottoman period in particular the dar was reduced to an informal composition of volumes centred on a void.74 The fact that there is a qa’a in every building aggregate from the beginning of the sixteenth century on confirms that it was the leading type in the typological processes of Cairo. We find it not only in palatine
Fig. 72 Dar Sarsiri plan and section north-south through the maq’ad (From B. Maury, Palais et maisons du Caire, IV, Cairo, 1983).
the building type but in middle-class buildings. It is especially distinctive in a certain type of collective residence known in Cairo as a rab. There were two types of collective residence in Ottoman Cairo: the hawsh and the rab.75 The first was a large enclosure, sometimes as large as an entire block, given over to the poor where they could put up some sort of shelter. More a shantytown than anything else, the building tissue was easily reabsorbed in more recent times and reapportioned without leaving any obvious traces.76 The second type was closer to the modern version of a middle-class residential apartment complex. It was formally structured and consisted of a series of apartments derived from the qa’a, distributed around a courtyard in duplexes or sometimes triplexes. Each apartment included a portion of the terrace, the only external outlet for the residence. Often a rab was next door to a wakala caravanserai, in which case the ground floor and mezzanine of the complex were used for storage and business transactions and the upper stories were purely residential. The apartments opened onto the landings of a series of stairways that led to the outside. In this way the autonomy of both areas was maintained. This layout is legible on the outside of the building through the surface punctured with windows and mashrabiyyas; it rested on a solid rustic stone base with a full-height monumental entrance cut into it. In addition to the distribution around a balcony and the maisonnette apartments that were so similar to the buildings of Le Corbusier and the hofs of Vienna, the clever use of space and the grouping of ventilation and sewerage pipes in the interior also helped to create a modern machinelike habitat. The hierarchy of the spaces inside each dwelling was clear and straightforward. A rectilinear sequence traversed the building unit: first the entrance or dilhiz, the latrine, and the stairs leading upstairs; then the main room made up of a durqa’a and iwan, the far wall of which was lined with windows. The kitchen was located on the mezzanine above and looked out onto the room below. An identical room, but more ordinary in height, was located on the uppermost level. It gave onto the terrace and was an exclusively female domain. Despite differences in nomenclature the ritual and sequence of use of the traditional qa’a were exactly the same, whether located in palaces or houses. More modest versions exist, like the Rab as-Sadat, that do not follow this tendency, however, unless the apartment, little more than a cell, is read as an extremely unspecialized form of the qa’a. Even if the Ottoman qa’a is reduced in overall size - the qa’a of the Dar Kathuda is less than 20 meters long- it is no less imposing a figure. Mashrabiyyas were used liberally within, often taking up a whole back wall. The iwan was fairly independent and was allowed to develop transverse axes tied to a window that faced outside. The meeting of Ottoman culture and local customs at the time of Muhammad Ali produced a new type, the fasha.77 A combination of a large central hall and grand staircase, it was used in large buildings and condominiums. The distributive hierarchy of such a large area began with a large courtyard from which a number of different stairways led to various fasha. These in turn acted as internal covered courtyards, onto which the
113
Fig. 72.a The wakala and rab’ of al-Saraibi. Axonometric section and plan of the main floors. The commercial area is less brilliantly organized; the residential cells are organized around only one balcony open to the courtyard.
114
after amnesia
the building type
115
Fig. 73 (Left The wakala and rab’ of Zulficar. Axonometric section and groundfloor plan. The plan is similar to a specific serial building that grew in phases like a residential block. Note the inconsistencies of the corner solution. To the left is the matrix tissue, then the stripes of two planned tissues perpendicular to the formal tissue, and later a closing as a tissue of connection. (Right) The figure shows in axonometric section and groundfloor plan the wakala and so-called Rab Bazaar in Cairo; one of the most monumental examples, it was probably bu lt at the end of the seventeenth century. On the third floor, the exterior corridor leads to nineteen apartments organized on three floors and varying in size between 700 and 1,400 square feet. The organization of the whole building favours social contact and creates a gradual transition from private to public space.
116
after amnesia
individual apartments open, and in which the common service areas were located. The layout of the apartment was similar to that of the rab: an elongated living room was divided into a vestibule, reminiscent of the durqa’a, and an iwan with a projecting balcony (balcon). The niches of the qa’a were large enough to function as secondary rooms. A lavish version of the type had a separate salun for visitors and an odt al-a’ad for family use; it was usually entered through a vestibule. Even the most modest versions included a living room and a separate service room. A detailed survey of modern Cairene apartment houses has been carried out by a group of French scholars led by Jean Depaule, and it is therefore not necessary to describe them here. We need only point out that, in spite of a visual language that placed the Cairene apartment in an international panorama, all these types originated and evolved in local typological processes.78 Imported fashion may cover, but cannot erase, this evolutionary process. Perhaps the most important lesson of the Cairene typological process is that it elaborates the parallels and the simultaneity in the progressive specialization and hierarchization of the urban tissue and the complex growth and changes in the house. In the modern history of Cairo, the passage from the first Fatimid settlement to the Mamluk city and the Ottoman metropolis is mirrored by the changes in the house: specialization and contraction of the courtyard house in the qa’a (a more compact residential type subject to insertion either in a house, or a palace, or a rab, or even a modern condominium), as well as the formation of collective residential types that often joined to make a special tissue. Why a similar phenomenon did not occur in the Maghreb is a puzzle. For example, in the two centuries preceding Western domination Tunis had a number of residents and a quality of services that could have been the envy of the East. Nevertheless, the residential tissue of Tunis is much less hierarchized, that is, less organic and complex, than in Cairo and the leading residential type less elaborate-so much so as to appear to be without variation in the tissue. Summing up the typological processes of the Mediterranean courtyard houses, two large families are recognizable, which correspond more or less to the division of the Mediterranean into east and west. On the one hand, the house of the Maghreb and Andalusia, which belongs to an area that employs mixed building technologies, and where the elementary cell, derived from the taddart and aggregated in domino fashion, strongly embodies the entire system. There the Wast ad dar is the centre of the composition, the light and air lung, which identifies with the house itself. On the other hand, all the houses of the east from Venice to Lebanon belong to a family whose distributive plan privileges tripartition and centrality. This similarity between Venice and the Ottoman world could be ascribed to the presence of a type of domus substratum, which under various guises would have remained a prisoner of the fabrics and building consciousness inspite of an excessive chronological difference.
Fig. 74 Cairo. A Fasha house ( From O.Blin. Le Caire XIX-XX siècles. De la Fasaha à la sala comme modèles, in Cahiers, op.cit., 1987, pag. 96)
the building type the colonial tenement house in algiers Colonial domination represents a caesura for the typological processes of the dominated countries equal to the Black Plague that struck Europe in 1348 and left cities without strength for several centuries. When colonialism finally leaves the colonies, the process begins again, because the cycle of life imposes the paradox that the built environment left by the colonial administration becomes in effect a cultural patrimony of those who were dominated. The building history of Algiers was the reverse of Cairo’s in the sense that a dominating foreign culture elaborated typological processes in synchrony, not with the local requirements, but with changes in the parent country. The common types that followed in Algiers are all apartment houses, produced in turn from the modification of the row house, imported prevalently from the south of France. They represent a clear break with the courtyard house, with which the qasbah continued to grow during the Ottoman period, before the conquest. The 130 years in which these typological processes took place are nevertheless interesting for the legibility of the phases of the leading types. Since the thesis we propose is that of the progressive alienation in the 19th century between type, building aggregation and urban fabric over time, which will then be codified by the avant-garde of the Modern Movement as a conquest, it is important not so much to enter the details of the infinite typological variants as it is to reconstruct the route and changes of the leading types.79 The building history of Algiers can be divided into four phases based on the seminal changes in building type and urban fabric. The first phase began with the French occupation in 1830 and lasted until 1846. During this first phase, even perhaps until 1854, building involved mainly the restructuring of the existing fabric. Despite appearances, French interventions reproduced the functional logic of the existing medina, reinterpreting it. A set of reconstruction techniques can be seen along the Rue Bab Azzoun, a new road arcaded on both sides that was carved into the existing fabric, but without major demolition; new facades were built for the affected structures. The relative flexibility of the existing courtyard house made this restructuring comparatively easy; the internal facades remained intact and preserved a considerable part of the original character of the type. Many trapezoidal plots containing smaller courtyards resulted from cutting out the original blocks, which were then covered and occupied by stairways. The courtyards in the plots on the Rue de Chartres, another restructured street, were small to the point of being uninhabitable. They rapidly filled up, turning the structure into stacked-up rooms without air or light. The situation was better in the Place de Chartres, a perfect square cut out of the fabric, surrounded by free-standing porticoes on three sides. The new urban quarters on the cliff along the edge of the sea were built on a regular grid of 20-25 m x 30-40 m. The rectangular plots were dominated by the apartment building type, organized in four rows with two or three bays, and an asymmetrical access and stairway. The topography often demanded simpler and more archaic solutions as well
117
118
after amnesia
Fig. 75 The scheme reconstructs the typological process of Algiers’s tenement houses (only the leading types are mentioned) beginning in 1830. The chronological function is on the X-axis divided into four phases. The levels of scale are on the y-axis and permit us to see the inter-scalar relations of the type with the urban implant, urban fabric, building tissue or block building type, facade and load-bearing structures.
as many synchronic variations, such as the three-module types found in Rampe Rovigo (now Debbih Cherif) or the redevelopment around rue de Tanger (now rue Chaib) called trois fenetres.80 The peculiar topographical conditions required a great variation in building types. In the area of rue d’Isly the inclined ground did not allow large blocks and thus led to the use of more serial synchronic variations in which the first two floors cut into the sloping ground. Both restructured and new blocks were built according to the French building regulations of 1784 which prescribed precise proportions between the width of the street and the height of the building, calculated at the cornice: 14.62 meters for a 9-m-wide street, and 17.54 m. for a 12-m-wide street.81 Masonry continued to be widely used. Masonry bearing walls were usually perpendicular to the direction of the street and were at a 3 to 3.5 meter distance - a dimension derived from local building practices, first found in the small Ottoman Turkish palaces in Algiers.82 The modular grid of the windows is the main characteristic of the facades; otherwise they
the building type
119 Fig. 76 Algiers. Place du Gouvernment (today Place des Martyrs). It was the first French intervention in the core of the lower medina.
Fig. 77 (Left)Tenement house in Place du Gouvernment built in the 1840s. Dimensions and proportions of the facade are based on 10-foot modules. (Right) Tenement house in Place Yacef. This house built in the 1850s shows a strong horizontal hierarchy: basement, mezzanine, two second floors, attic, and a vertical division of rows separated by pilasters. The balconies, according to tradition, are located on the third and fifth floors.
120
after amnesia
are almost without any decoration or articulation, but the even number of bays creates a non-symmetrical access. The only hierarchy is created by the strong differentiation between the porticoed base and the rest of the facade. The roof is generally pitched or contains an attic. The first phase of building in colonial Algiers demonstrates that in spite of the importation of alien ideas and technology a certain local substratum remained active, as can be seen in the measurements of the ceiling spans, the reduced size of the urban blocks, and in general the tendency to maximize the existing fabric. The second phase, from 1854 to 1881, coincides with the Haussmanian transformation of Paris and is characterized by a new wave and new type of building activity, especially in the Mustapha area. The radial plan, as a general characteristic, does not derive from the urban block, but from a geometrical abstraction. Large triangular shapes- unusual in the past started to appear, and were generally occupied by small shops located at the corners. The plots were divided according to the bisection of the angle rule and took trapezoidal or triangular shapes; but it was not the shape so much as the size of the plot that was important; it definitely increased from the preceding phase.83 The width of the street was determined using Paris’s new building regulations of 1859, which favoured the 45-degree diagonal for grand boulevards wider than 20 m. In the rectangular blocks around the Place de la Republique, in the vicinity of the Palais de Justice, along Boulevard Zirout Youcef, and other locations, L- or T-shaped plots were now most common. Soon thereafter deeper plots with synchronic variations (such as blocks constituting only a single plot) appeared for the first time. In the interior distribution of the apartments in these plots arbitrary complications were introduced, such as aligning the apartments at 45 degrees to the external walls and multiple stairwells - as if the new type aimed at recuperating in a single plot the lost complexity of the old medieval fabric. In the triangular plots, the axis of symmetry was established by a cut corner, later replaced by bay windows or balconies. The cut corner also played an important role in the rectangular plots as well, as in the case of the Place de la Republique and, especially, the rue Boumendjal Ali. The composition of the facade shows the axis of symmetry, and an odd number of bays that led to the central location of the portal, and the division into two equivalent parts; significantly different from the facades of the earlier phase, which had an even number of bays and thereby lacked a central axis of symmetry.84 At the same time the building type started to detach itself from the urban block. Since it was not advisable to divide the central room in half just to accommodate the axis of symmetry, this room was indifferently assigned to one or the other apartment on each floor, thereby creating differences in size on each level that were not perceivable from the outside. The horizontal hierarchy of the facades also became stronger. The moulding strongly separated the mezzanine from the rest of the facade, now even more articulated by the presence of rows of balconies with wrought-iron parapets usually on the third and the fifth floor.85 Decoration became heavier with the use of the giant order, cornices, and caryatids
Fig. 78 (Above) Tenement house Cognon on Rue Dumont d’Urville, built in the 1980s. The corner apartment is centred on the corner rotunda, which acts like a hinge. (Below) Tenement house on Rue de Constantine, built in 1905. Modules of the space. Second floor and attic plan. The triangular form of the plot induces a difficult distribution of the internal part that appears as an encroachment tissue. Main rooms are arranged along two rows parallel to the front and served by two long corridors. Note the difficult correspondence between the layout of the windows and the layout of the rooms to the right. The building type is reduced to a mere surface to be cut up.
the building type that emphasize the rows, creating for the first time very articulated facades. Architectural styles were mixed with great ease, from the Greek-Roman and the Renaissance, to French architecture of the seventeenth and the eighteenth centuries. The third phase, from 1881 to 1915, was marked by the winding developments on the hillsides, presenting an articulated path system to connect different altitudes. Diagonal cuts are frequent - Rue Addoun Ahmed, Rue Arezki Hamani, Rue Ferroukhi Moustapha, and Boulevard Mohamad V - and generated large triangular and trapezoidal plots. They became even larger, partly due to the financial organization of contractors and site work. The actions on different scales became progressively uncoordinated: the city plan laid out in a network of curvilinear streets was separated from the urban block, the block in turn was separated from the residential type. The prevalence of the triangular and trapezoidal plots, dictated by a new urban design paradigm of a radial street pattern, problematized the internal distribution of the type. The architects compensated for the separation of type from the building tissue by a richer treatment of the facades, employing symmetry, horizontal and vertical mouldings, cornices, and other decorative elements. The building type was characterized by two external bays of aligned rooms on the two opposite facades, and two internal bays of secondary rooms lighted and ventilated through the central bay which includes light-wells and stairways. One of the more interesting synchronic variations of this period is at the intersection between Rue Chaib Ahmed and Rue Boumenjel: a trapezoidal plot with a small central courtyard covered by a skylight. A complex vertical system of cantilevering ramps leads to the second floor, from which normal stairways ascend to the upper floors. On each landing there are two apartments with rooms distributed along a corridor, favouring the location of the largest apartments on the street side. The extremely rich decoration -statues, mouldings, lights - suggests the high bourgeois origin of the apartments. In this period there are frequently buildings that are specialized, containing two levels of commercialization. In fact, the large hall is moved to the piano nobile, while retail business takes place on the ground floor, and some ateliers are located at the mezzanine level.86 The French building regulations of 1884 were now enforced; without changing the maximum height of the cornice, they allowed two additional floors by virtue of a much steeper pitched roof.87 The new regulations also allowed cantilevering out from the facade. The cut corners were replaced by rotundas or bay windows (corresponding to the most important room in the apartment) which give the facades a plastic look, while simultaneously increasing the view onto the street and the area of the apartments. Very long facades, over 200 meters (over eight times the length of the blocks in the first phase) were designed, emphasizing hierarchy in the composition to avoid the monotony endemic to such long facades. Rectangular bay windows were introduced in the middle of the facade to break the surface and create a rhythm and vertical hierarchy. This virtuosity in facade treatment obviously had little relationship with the internal
121
Fig. 79 The corner of the same bu lding in Rue de Constantine deftly solves the problem of the difference of level of the two streets.
122
after amnesia
layout of the apartment, and from the outside it is not possible to recognize the distribution of the apartments, or their rooms or functions. There are innumerable examples of these blocks; it would be enough to mention the apartment block near the Saint Augustin church, and the apartment block at the intersection between Boulevard Khemisti and rue Emir el-Khettabi. The horizontal hierarchy is now deemphasized; all floors are the same height, and a strong vertical modulation made of pilasters (parastas) and sculptural decorations, statues, caryatids, begin to appear. After 1884, the simple grid of the openings is not used anymore, and after 1902 we can observe a modular division of the windows, now decoratively protruding out of the facade- a harbinger of Art Nouveau. Toward the end of the nineteenth century the practice of borrowing stylistic elements from the “Arab” architectural language increased and was freely mixed with the colonial vocabulary. Discrete and yet interesting designs arose from this eclecticism, especially in corner solutions: two examples are the corner building at the intersection between rue Ben Mhidi and Boulevard Ben Boulaid, and the Garcia building in front of the Aletti Hotel.88 In the third phase, type was reduced to a single surface, which was also cut randomly to adjust to the plot. Larger apartments appeared, occupying the entire depth of the block with up to five bays. The articulation of the long facade was achieved by alternating bay windows and recesses and using very elaborate round corners. In the last phase, which continued until 1962, the new development in the periphery was based more and more on topography, with curvilinear paths and irregular plots like large built-up islands, making it more and
Fig. 80 The side facade along Rue d’Isly of the same building. Note the hierarchy obtained by the recessions and protruding of the bay windows.
the building type
123
Fig. 81 The tenement house Garcia built in the 1920s. Note the rotunda in the corner in “Arab” style. Moorish is what the French believed was what the Arabs should think about Arab architecture.
more difficult to construct buildings parallel to the street and leading to the detachment of the apartment building from the block. Every building undertaken now took the new independence as a chance to differentiate itself from the others, starting a never-ending game that led only to the banal. After the 1930s, the new designs followed the paradigms of the Modern Movement, that is, the notion of “the platonic geometric solids in the garden,” and denied the dictates of the terrain or continuity with the existing fabric, as in the Aerohabitat project by Miquel and Bourlier and the quarter in Champ de Maneuvres by Zehrfuss. The separation between building type, building tissue, and urban fabric is complete. As a synthesis: we have found the elementary type, the initial matrix of the processes of the house, in the elementary cell. This last can evolve through successive duplications. The evolutionary process of changing the row house into apartment house was not considered here; though it is important to justify the creation of the apartment, the row house as a phenomenon is limited to the northern European part of the Mediterranean that is marginal to the Arab Islamic world.
124
after amnesia
The courtyard house, on the contrary, requires considerable attention because it is the residential type of choice in the Arab Mediterranean. The figures showing its evolution tried to cover all cases. Of particular significance are the two processes of commercial specialization (the taberna process) and subdivision and fragmentation (insula process), which are at the root of any phenomenon of the metabolization of the Arab fabric. The moments of reading at the level of the building start from the consciousness of the building organism (excluding special buildings). The articulation of the organism in systems of structures follows: the wast-ad-dar of the Maghrebian house has been used as an example on account of its powerful ability to qualify as the house globally. After that we pass on to structures and finally to elements. To give examples of all the Mediterranean residential processes is not only impossible but of little interest. An encyclopaedic treatment of the topic would only distract attention from the real goal, which is to introduce some of the main guiding principles that teach one how to read the urban landscape. With regard to different cultures, some characteristics, such as the similarity of houses in Aleppo, Lebanon, and Venice with the Ottoman sofa house, do not represent a phenomenon of cultural osmosis that could have taken place under the Ottoman Empire, but a phenomenon that has its roots in the Levant, in the basic type or substratum of the domus. The opposite cases of Cairo and Algiers from the first half of the nineteenth century show two different attitudes toward modernism: in the case of Cairo, a transition without shocks as the qa’a slowly evolves into the vestibule of the modern apartment; in the case of Algiers, an alien European culture violently combines with a local one to produce a radical choice in favour of building types (the apartment) and aggregations (blocks) imported from abroad. These types have their own history up to the war of liberation (19541962). A closer look shows that not even in Algiers did the local process completely give way; surveys show that periodically it reappears on the surface. Therefore Algiers has its own specific process that is a synthesis of two traditions and not a simple cloning of the French type. From Algiers we can learn an important lesson: to look with serenity on architecture, free of ideological prejudices. Architecture is the product of the civilization that created it and the (sometimes different) people who have lived in it and who exercise the right to adapt it to their customs and the duty of considering it as their own historic heritage.
125 notes 1. G. Cataldi, F. Farneti, R. Larco, F. Pellegrino, P. Tamburini, Tipologie primitive. I “ tipi radice “, (Florence, 1982), and G. Cataldi. Le ragioni..., op.cit., pp. 91-106. On trulli, see E. Allen, Stone Shelters (Cambridge, MA, 1969) and G. Simoncini, Architettura contadina in Puglia (Genoa, 1960). On the tower typologies of Yemen, see R. Lewcock and R. B. Serjeant, San’a, An Arabian Islamic City (London, 1983), especially the chapter “Houses of San’a,” pp. 436- 501. 2. The elementary cell is constant in the great serial buildings, of which it comprises the measurement module, where the unity of the organism is guaranteed by the hierarchization of the spaces, like the rab, or the Mameluk and Ottoman buildings in Cairo. The unusual spaces of the latter-halls, double-height spaces inside-are submultiples of the cell, whose presence guarantees coherence with human scale. Not even religious buildings, which expand interior space until it adjusts to the collective scale, lose human scale, since they contain a reference to the elementary building dimension; in the Arab mosque the outsized space of the courtyard (sahn) is brought back to human scale by the portico and the facade of the prayer hall, onto which is projected the row of interior bays, or in the courtyard of the Great Mosque of Cordoba by a modular planting of orange trees aligned with the aisles; in the Ottoman mosque, where widening is an intentional effect, the wisdom of interrupting the greater order with a balcony obtains the same effect. 3. The axonometric projections from below developed by Choisy - a threedimensional representation plus plan - were often limited to a single span to represent a whole building (A. Choisy, Histoire de l’architecture, 2 vols. (Paris, 1964). 4. The inverse process is also possible, however. The economic and military power of a city can exert strong cultural pressure on the countryside and influence its processes. This would explain the opulence and distributive articulation of rural towns in Syria’s limestone mountain range during the Byzantine period. An excavation at Belyounech in Morocco uncovered a rural housing complex connected to a sophisticated hydraulic network. The houses are quite urban in nature and resemble some of the Alhambra’s palatial structures.
They are most likely the country houses of wealthy city dwellers. See P. Cressier, J. Hassar-Benslimane, and A. A. Touri, “The Marinid Gardens of Belyounesh,” The Garden as a City and the City as a Garden, in A. Petruccioli, ed., Environmental Design 1 (1986), pp. 53-57. 5. W. Giese, “Los tipos de casa de la Peninsula Iberica,” Revista de dialectologia y tradiciones populares 7 (1951), pp. 582-583. 6. M. C. Delaigue, “Ethnoarchéologie de l’habitat traditionnel à toits plats dans l’Alpujarra rural,” Ph.D. diss., University of Lyon, 1986, pp. 83-86. 7. A. Bazzana, Maisons d’al-Andalus. Habitat médiéval et structures du peuplement dans l’Espagne orientale (Madrid, 1992), p. 169. For a more general discussion, see C. Flores, Arquitectura espagnola popular, 5 vols. (Madrid, 1973). 8. F. Ragette, Traditional Domestic Architecture of the Arab Region, Stuttgart, 2003, p. 112. See also A. Adam, La maison et le village dans quelques tribus de l’Anti-Atlas (Paris, 1951). 9. F. Ragette, Architecture in Lebanon.The Lebanese House during the 18th and the 19th centuries (New York, 1980), pp. 116ff. 10. For construction, we proceed with the four walls with arches that contain an accommodation for the ridge centering, and then the groin centering is placed with the lowest springer. The whole is covered with irregular shafts and afterwards with a matting surface. This is then finished off with a mud cover surface that makes the covering plastic and waterproof. The result is a visible and very serial cross vault. For a closer examination, see S. al-Aamiry and J. Cejka, “Das Palaestinen-sische Haus,” in Pracht und Geheimnis, exhibition catalogue, Joest, 1987. On the aggregations of the Palestinian cell and the study of the houses, see P. Revault, S. Santelli and C. WeillRochant, Maisons de Bethlehem, Paris, 1997.
“Una mezguita califal en las dunas de Guardamar del Segura,” Archeologia Medieval espagnola, vol. 3 (1986), pp. 505-520, and the figure on p. 516. 13. F. Braudel, The Mediterranean and the Mediterranean World in the Age of Phillip II (Berkeley-London, 1995). 14.V. Carette, Etude sur la Kabylie (Paris, 1848); A. Hanoteau and A. Letourneux, La Kabylie et les coutumes kabyles (Paris, 1893). 15. Recently I had the opportunity to visit the Semiramis project in the Carrieres Centrales quarter of Casablanca, designed by Candilis, Josic and Woods as low-cost housing. Each apartment originally included an open loggia that was subsequently transformed by the occupants to suit their specific needs. An extraordinary variety of typological creations resulted, most inconceivable from the point of view of the designers. See G. Candilis, A. Josic, and S. Woods, A Decade of Architecture and Urban Design (Stuttgart, 1968). 16. J. Berque, Les structures sociales du Haut Atlas (Paris, 1955); C.Vincente, “La maison kabyle,” Cahiers des Arts et Techniques de l’Afrique du Nord 5 (1959); R. Basagana and A. Sayad, Habitat traditionnel et structures familiales en Kabylie (Algiers, 1974). 17. The most extreme interpretation of the Kabyl house with a strong accent on the symbolic meaning of the components is found in P. Bourdieu, La maison ou le monde renversé (Geneva, 1972), p. 19.
11. E. Riza and P. Thomo, “Albania,” Architecture traditionelle des pays balkaniques (Athens, 1994), p. 44.
18. In the following passage Vittorio Gregotti underscores the tectonic importance of the courtyard as “an architectural act par excellence, the enclosure not only establishes a specific relationship with a specific place but is the principle by which a human group states its very relationship with nature and the cosmos. In addition, the enclosure is the form of the thing; how it presents itself to the outside world; how it reveals itself.”V. Gregotti, “Editoriale,” Rassegna 1 (December 1979), p. 6.
12. An unusual case of side-by-side cells located along the short side forming a thread-like tissue is found in a medieval structure at Guardamar de Segura in Spain. There is a niche in each cell that was probably a mihrab. Could this have been a village exclusively of chapels? R. R. Azzuar,
19. J. E. Campo, The Other Sides of Paradise: Explorations into the Religious Meanings of Domestic Space in Islam (Columbia, South Carolina, 1991); P. Oliver, Dwellings:The House Across the World (Austin, Texas, 1987); G. T. Petherbridge, “The House and Society,” in G. Michell, ed., Architecture of the
126
after amnesia
Islamic World: Its History and Social Meaning (London, 1978), pp. 193-204. 20. G. Semper, The Four Elements of Architecture and Other Writings (Cambridge, 1988). 21. G. Buti, La casa degli Indo-europei. Tradizione e archeologia (Florence, 1962). 22. On the continuity of the residential types, there is no substantial difference in Carthage between Byzantine houses and the early Arab houses. See M. Fantar, “Présence punique et romaine à Tunis et dans ses environs immédiates,” in Antiquités africaines, 14, 1979. The claims made by theorists of sustainable technology that the courtyard is capable of thermo-regulation are less than convincing. G. Scudo, “Climatic Design in the Arab Courtyard House,” Technology. From Tradition to Innovation, ed. A. Petruccioli, special issue of Environmental Design, 1988, no. 1, pp. 82-91. 23. Why the Yemenite tribes, which represent the most cultured part from the building point of view of the army of the Caliph, did not bring with them a type not only deeply rooted in their own housing culture, but relatively evolved, remains a mystery. Certainly at the time of the Prophet, the tower house could not have the technological development that we know, and it was instead lower, at most three storeys. The only explanation lies in the superiority of the courtyard type with regard to aggregability and distributive qualities. The tower house, in fact, comes into being isolated in the Yemen countryside, and is aggregated with some difficulty into rows in urban fabrics, but above all many problems are created by the fact that the gynaeceum, located on an intermediate floor, can be passed through by outsiders. Of the vast bibliography on the Yemenite house and the Arabian peninsula, we single out S. and M. Hirschi, Architecture au Yemen du Nord (Paris, 1983) and M. Nicoletti, Architettura e paesaggio nello Yemen del Nord (Bari, 1985). See also S. Nilsson, “Notes on the Vernacular Architecture of the Ashir,” in A. Petruccioli, ed., Environmental Design 1-2/ 1995. For a hypothetical reconstruction of the Sana’s house at the time of the Prophet, see R. Lewcock and R. B. Serjeant, Sana’a an Arabian Islamic City, op. cit., pp. 436 ff. 24. G. Petherbridge, “The House and Society,” ibid., p. 176.
25. A. Bazzana, pp. 169 ff. 26. N. J. Habraken, The Structure of the Ordinary (Cambridge, MA, 1998), pp. 146149. 27. On the Maghrebine house see: Cahiers des Arts et Techniques d’Afrique du Nord (Toulouse, 1959). On the domus, see K. Lehman Hartleben, Denkmaler antiker Architektur (Berlin- Leipzig, 1936). A reinterpretation of the role of the Roman patio house and its relative fabrics is in G. Caniggia, Lettura di una città: Como (Rome, 1963). 28. On the row house, in addition to the cited studies by G. Caniggia, see G. L. Maffei, La casa fiorentina nella storia della città (Venice, 1990). 29. J. Revault, Palais et demeures de Tunis XVIIIe et XIXe siècles (Paris, 1971). 30. More precisely, 9 m and 18 m facades are found at Priene in two different quarters; 15 m at Olinto and Kassope; 18 m at Dura Europos, and even 21 m at Abdera. The oversized facade of the last example can be explained by the fact that the courtyards there line up with the predominant side parallel to the route onto which the entrances open. W. Hoepfner and E. L. Schwandner, Haus und Stadt im Klassichen Griechland (Munich, 1994). 31. On Roman metrology, see A. Martini, Manuale di metrologia (Turin, 1883). 32. A non-commercial use derived from the taberna process is found in the Maghreb in the annex to the dar called an ali in Tunisia and Algeria and a masriya in Morocco. Connected to the upper story of a house, its windows face outward; it is reached from a staircase either outside or in the vestibule of the dar. It is usually rented out, so it represents both an investment for the family and a means by which the city could absorb population increases without having to create an ad hoc leading type. It has played an important role in modernization and the creation of multi-family dwellings by allowing the dar to rise vertically and by turning the house so it faced outward. Regarding the historical importance of the masriya to the Maghrebian tissue, Jean Claude Garcin notes that, according to a census at the time of the Almohad Caliph al-Nasr (1199-1214), there were 89,236 dars and 19,041 masriyas in Fez. According to al-Maqqari the term masriya in eleventh-
century Cordova referred to rooms in a house that were rented out. J. C. Garcin, “Quelques questions sur l’évolution de l’habitat médiéval dans les pays musulmans de la Méditerranée,” L’habitat traditionnel …2, pp. 369-385. For descriptions of masriya, see J. Revault, “Le Dar Caid Bel-Hassen et le Dar Zouiten à Fes,” L’habitat traditionnel … 1, pp. 275-298. 33. A. Giannini, “Ostia,” Quaderno dell’Istituto di Elementi di Architettura e Rilievo dei Monumenti vol. 4, pp. 36 ff. 34. Alexandre Lezine notes that in a Tunisian block on rue de Tamis, three courtyard houses were transformed in this way after the seventeenth century. See A. Lezine, Deux villes d’Ifriqiya, Etude d’archéologie, d’urbanisme et démographie (Paris, 1971), p. 160. When this development became too widespread it altered the urban morphology: for example, in the qaissariya of Fez the entire northern residential tissue has been assaulted and corroded by commercial activity. The appearance of the fabric of Fez el Bali is deceiving: to the living suq is opposed the interior of the fabric with many houses reduced to their outer walls, all those houses that have no outside outlet. 35. Some classical archaeologists call any tall building an insula in order to distinguish it from the domus; others use the term to mean a block, a usage which is justified by the fact that the estimate of Rome’s population in its apogee is based on the calculation of the number of insulae. In Rome a 17.7 meter height limit was established by Trajan; see A. Chastagnol, La préfecture urbaine à Rome sous le Bas Empire (Paris, 1960), pp. 369-371. In fourth-century Constantinople a height limit of 29.5 meters was set by the code of Justinian (VIII: 10, 12,14). On the insula of Ostia Antica, in addition to the previously cited Giannini essay, see J. E. Parker, “The Insula of Imperial Ostia,” Memoirs of the American Academy in Rome vol. 31 (1971). On the insula of Rome, see G. Lugli, Roma antica. Il centro monumentale (Rome, 1946), and G. Calza, La preminenza dell’insula nella edilizia romana (Rome, 1915). 36. The balcony in a courtyard that has undergone the insula process gives the maximum result insofar as it allows a single staircase to provide access to many residences. Conceptually it is equivalent to the loggia, which limits itself to the size of a passageway (approximately 4 ft. or 1.2 m.)
127
notes
and runs along the entire interior facade of the building on every story. The local version of the courtyard type in Milan is known as a balcony-house (casa a ringhiera) and it later appeared in the Modernist movement. Candilis, Josic, and Woods have used balconies as corridors for distribution in various projects for subsidized housing in Morocco (the Semiramis Block) and Algeria (Tersade Block). 37. In reality, the passage from the courtyard house to the linear apartment house is part of a natural trend to respond to densification, but here we are speaking of demolition and the substitution with a building of much greater mass. On typological substitutions and their corresponding social features in old Cairo, see L. Christians, O. Greger, F. Steinberg, Architektur und Stadtgestalt in Kairo (Berlin, 1987), especially the section on the Husayniyya quarter, pp. 84 ff.
the access axis of the courtyard. In addition to this example, the distributive element is also apparently found in Semitic regions.
G. L. Maffei, Composizione architettonica e tipologia edilizia. Il progetto nell’edilizia di base (Venice, 1985), p. 160.
42. R. Riche, “La maison de l’Aurès,” Cahiers des arts et techniques d’Afrique du nord, vol. 5, pp. 30-36 and A. Petruccioli, S. Miller and M. Bertagnin Inscribing Minority Space in the Islamic City.The Jewish Quarter of Fez (1438-1912), in JSAH Journal of the Society of the Architectural Historians, 2001, pp. 308-327.
48. C. Cambazard - Amahn. Le decor sur bois dans l’architecture de Fes, ( Paris, 1989).
43. Oliver Aurenche. La maison orientale. L’architecture du Proche Orient dès origines au milieu du quatrième millenaire, 3 vols. (Paris, 1981). 44. F. Ragette, Traditional Domestic Architecture….., op. cit., p. 24. There is no contradiction between the two interpretations since the section of the taddart depends on the available wood materials available.
38. In Algiers a metal screen is ingeniously placed over the courtyard to act as both a surface to be walked on and a covering with a cloth in case of rain. This creates two domus, one on top of the other, while still keeping the courtyard as a global space inside the house; a sort of insula is therefore created.
45. A. Touri, M.A. Hassaui e G.C. Barbato. Le projet pilote de restauration et réhabilitation du Palais Dar Adiyel à Fes (Rome, 1999).
39. A question that arises and remains unresolved is that of family structure. The dar, especially in old cities, was associated with the extended family - a structure closer to the Latin idea of gens than to the modern nuclear family. The growth of a family generates new units, often producing building complexes. It is not unusual to find that over time an extended family grew so numerous as to generate its own neighbourhood. See figure 38, a survey of an Essaouira house on the rue de la Mellah by M. Accorsi and A. Petruccioli. Also M. Cote, L’Algerie ou l’espace retourné (Algiers, 1993), especially the figure on p. 255: the transformation of a patio into a route that runs all the way to the rooms.
47. The ceilings in Essaouira are very different from those in Fez. The beams are of thuja rather than rare wood, and are no bigger than 3.2 meters. They support a secondary structure covered with a layer of tassuit branches. The floor is made by covering a layer of dry leaves with and a mixture of mortar and earth. The typological specificity of each area is confirmed by the fact that seemingly similar ceilings are in fact realized according to quite opposite concepts. The Neapolitan ceiling is described by Caniggia as follows: “In Naples … a tradition of structures vaulted in light stone (pumice, lapillus) results in an apparently wooden ceiling being constructed with small barely stripped beams … a ceiling which in and of itself does not support … it could do so statically if isolated. In reality it is not so much the wood frame that provides support as the mass of lapillus mixed with mortar that thickly covers it (up to a meter in places). The wood frame acts as a disposable mould for what is laid upon it: it is the authentic structure insofar as it forms a sort of natural vault resulting from the settling of the casting material when the lens is separated from it.” G. Caniggia and
40. A detailed description of the domestic architecture of Fez can be found in the exhaustive work of Jacques Revault: J. Revault, L. Golvin, A. Amahan, Palais et demeures de Fes: 1. Epoques merinide et saadienne (XIVe-XVIIe siècles), (Paris, 1985). 41. The archetype for the skifa is found in the so-called Door of the Spirits in the Chinese house; a wall recessed with respect to the external wall and perpendicular to
46. A. Grillo, “Traditional Building Techniques in Fes”, in A. Petruccioli ed., Technology from Tradition to Innovation, Environmental Design, 1/ 1988, pp. 38-47.
49. Some of the different patterns found in bricklaying include curtain, cross, oblique, herringbone, double herringbone, and stone and wood inlays. “Schema directeur d’urbanisme de la ville de Fes,” dossier technique IV, 2, in Les techniques traditionnelles de l’architecture et du decor à Fes (Paris, 1980). 50. S. H. Eldem seems to suggest that the cardak is a recent conquest where it categorically affirms that the historical typologies do not face the street. See S. H. Eldem, Turk Evi Osmanli Donemi, vol. III, Turkiya Anit Cevre Degerlerini Koruma Vakfi (Istanbul, 1987). 51.Various authors, “The Ottoman House”, Papers from the Amasya Symposium, 24-27 September 1996, BIAA monograph 26, 1998. 52. J. Bing and J.B. Harrington, “A Study of Words and Buildings-The Cardak of Former Yugoslavia,” in Environmental Design, n. 1-2, 1994-1995. See also D. Kuban, The Turkish Hayat House (Istanbul, 1995). 53. The extension of the surface of the patios was calculated by J.- C. David (in J.-C. David, Le paysage urbain d’Alep. Etude de géographie urbaine, op.cit, pp 168- 177) in a percentage that goes from 30% to 40% of the built surface of the lot and that is directly proportional to the extension of the lot itself. On the Aleppian house see: J.-C. David. “ Alep, dégradation et tentatives actuelles de réadaptation des structures urbaines traditionelles”, in Bulletin d’études orientales de l’Institut Français de Damas, Tome XXVIII ( Damaskus, 1975) and do. Syrie: systèmes de distribution des espaces dans la maison traditionnelle d’Alep, in Cahiers de la recherche architecturale, 20-21 ( 1982), pp38-47. A. Gangler Ein traditionelles Wohnviertel in Nordosten der Altstadt von Aleppo in Nordsyrien ( Tübingen, 1993) 54. Little is known about the domestic architecture of the Maghreb in particular. Recent ethno-archaeological studies such as E. Laoust’s work on central Morocco have allowed us to isolate the taddart, a long narrow cell that is fairly consistent from tribe to tribe. In our opinion the
128
after amnesia
taddart is the elementary type that in the Maghreb becomes a component structure, the bayt- on a par with the Easternimported T-shaped room - of both rural and urban homes. As the dwelling of transients, the taddart is nothing more than the stone version of a tent, placed along the perimeter of a courtyard, just as nomads would place their tents around the douar leaving room in the middle for their animals. E. Laoust, L’habitation chez les transhumants du Maroc central (Paris, 1955). 55. S. Muratori, Per una operante storia urbana di Venezia (Rome, 1959); P. Maretto, La casa veneziana nella storia della città dalle origini all’ottocento (Venice, 1986); G. Caniggia, “La casa e la città nei primi secoli,” in P. Maretto, La casa veneziana…, pp. 3-53. We learn something very valuable from these pioneering studies: that a non-critical consideration of partial historical facts in the reading phase can sidetrack a study at the moment of synthesis. After writing about Rome for twenty years, Caniggia turned again to Venice in the 1980s. He showed how the city, which Muratori presumed to be Byzantine on the basis of historical-literary data, was in reality of Roman origin. In addition, it was not simply a temporary settlement for refugees from Aquilea, but an established settlement of a relocated rural population who were dependent on the lagoon economy. 56. Part of Caniggia’s complex reconstruction of the typological process is summarized in figures 62-63; I have introduced some variations to simplify the various passages. A synchronic variant of early building with the room built transversely in the middle of enclosure generates a fairly diffused thread beginning with the 13th century, of which the Goldoni House in San Polo is a noble example. 57. P. Maretto, L’edilizia gotica veneziana (Rome, 1961). 58. M. Maretto, “Urban Morphology Basis for Urban Design: the Project for the Isola dei Cantieri in Chioggia,” in Urban Morphology, 1, vol. 9, 2005. 59. F. Ragette, Architecture in Lebanon, pp. 25-27. 60. H. De Beyle, L’habitation byzantine (Paris, 1903); A. Deroko, “Quelques réflections sur l’aspect de la maison byzantine,” Actes du Congrès international
d’études byzantines, 1955 (Istanbul,1955), pp.124-25. 61. There is considerable debate over whether the iwan is the product of an autonomous evolutionary process - a local interpretation of a pre-Islamic idea familiar throughout the Orient (the hypothesis I favour) - or simply an imported model. A Persian term, iwan (liwan in Arabic), describes a space -dependent or independent - closed on three sides by walls and completely open on the fourth. It is usually vaulted. The form can be applied to a platform or an entire building. For a detailed discussion of the monumental iwan, s.v.”Iwan” by O. Grabar in the Encyclopaedia of Islam, 2d. ed.
69. Ibid., pp. 276-279. 70. K. A. C. Creswell, The Muslim Architecture of Egypt, 2 vols. (Oxford, 1959), 2: p. 208. 71. That Santa Maria in Antiqua (located in the Roman Forum) is directly derived from a domus is evident when comparing the planimetry of the church with that of the so-called Surgeon’s House in Pompeii. Similar analogies exist throughout the Mediterranean culture area, particularly in Syria. 72. J. Revault, “L’architecture domestique du Caire à l’epoque mamelouke XIIIXVI siècles,” in Palais et maisons du Caire, 1. L’époque mamelouke” ed. J. C. Garcin,
62. J-Cl. David, Syrie: systèmes….., op.cit., p. 38.
B. Maury, J. Revault, and M. Zakariya (Paris, 1982), pp. 108-109.
63. M. M. Cerasi, La città del Levante (Milan, 1986), pp. 176 ff., and D. Kuban. The Hayat House (Istanbul, 1995).
73. Both André Raymond and Nelly Hanna discuss the important consequences this has for the urban whole. N. Hanna, “La maison Waqf Radwan au Caire,” L’habitat traditionnel ..1: pp. 61-77.
64. S. H. Eldem, Turkish House: Ottoman Period, vol. 1 (Istanbul, 1984). Ibid., pp. 26-31, for the planimetric drawings, in part reproduced in figure 67. Several scholars have noted that Eldem’s plans, more than reconstructions of a typological process, seem to be schematizations of real examples, taken from noble examples from the Marmara Sea region, with no distinction between common type and typological variants, listed from the simplest to the most complex. 65. D. Orhun. “ Is it the Turkish House or the Living Integrated House?”, in 7 Centuries of Ottoman Architecture. “A SupraNational Heritage” ( Istanbul, 1999), 276-284 66. Eldem vehemently rejects the idea of a marked Byzantine influence, arguing that no trace of the Byzantine house exists in Istanbul. Nevertheless the similarity with the Byzantine type called triklinium is striking. We can also note how the Venetian houses of Cyprus, an island in contact with Anatolia, do not have formal affinities with the Anatolian Ottoman house. 67. G. T. Scanlon, “Fustat Expedition: Preliminary Report 1965,” Journal of the American Research Center in Egypt 5 (1966). 68. A. B. Bey and A. Gabriel, Les fouilles d’al Foustat et les origines de la maison arabe en Egypte (Paris, 1921).
74. D. Behrens-Abouseif, “Alternatives to Cadaster Maps for the Study of Islamic Cities,” Urban Morphogenesis, a special volume of Environmental Design, ed. A. Petruccioli, 1993, nos 1-2: 92-95; and idem, “Notes sur la function de la cour dans la maison moyenne du Caire ottoman,” L’habitat traditionnel, 2: pp. 411-418. 75. In Grandes villes arabes, pp. 323-326, Raymond describes the hawsh and its presence throughout the Ottoman Empire. According to Clerget (op. cit., 1,p. 312) each hawsh in Cairo was village-like and housed thirty or forty families. There were many hawsh in the city and suburbs of Aleppo according to the traveller A. Russell, who describes them as “a wide space surrounded by a certain number of low, mediocre dwellings of two or three rooms each. The common area is haphazardly paved, with the exception of the space in front of each house...There are no fountains but there are numerous wells.” (A. Russell, The Natural History of Aleppo, vol. 2 (London, 1794), p. 36. 76. Jean Claude David located what is probably of series of Ottoman-period hawsh in the Qarleq quarter of northeastern Aleppo; J. C. David, “La formation du tissu de la ville arabo-islamique: apporte de l’étude des plans cadastraux d’Alep,” Urban Morphogenesis, pp. 138-155.
129
notes
77. M.Volait. “Grandes demeures du Caire au siècle passé”, in Les cahiers de la recherche architecturale, 20/21, pp. 84 ff. 78. Olivier Blin, “Le Caire XIX-XXe siècles. De la fasaha à la sala comme modèles,” Les cahiers de la recherche architecturale 20-21 (1987), pp. 96 ff. 79. A. Petruccioli. “Alger 1830-1930. Pour une lecture typologique des immeubles d’habitation”, in Algèrie. Mémoire et architecture, A. Petruccioli, ed., Environmental Design, 1-2/1992, pp. 104107. 80. This building type is directly derived from a type much used in Marseilles in the preceding century; see J. L. Bonillo, Type urbain et types domestiques: analyse architecturale des trois fenetres marseillaises (Marseilles, 1978). 81. L. Hautecoeur, “Le permis de construire,” in Académie des Beaux Arts, 1957-58; for building regulations, see Recueil des Lettres Patentes, Ordonnances Royales, Décrets et Arretes prefecturaux concernant les voies publiques de la Ville de Paris, ed. Deville and Hochereau (Paris, 1960). 82. Some of the ground plans and sections are published in Lucien Golvin, “Demeures d’Alger,” in L’Habitat traditionnel dans les pays musulmans autour de la Méditerranée, vol. 1: L’Héritage architectural: formes et fonctions, (Cairo: Institut Français d’archéologie orientale, n.d.), vol. 1,1, pp. 181-198. 83. See J. Castex, J. Ch. Depaule, Ph. Panerai, Formes urbaines: de l’ilot à la barre (Paris, 1980): “The dimensions of the triangular plot - the most frequent ones, indeed - sensibly vary and seem to exclude an optimum subdivision which would have been valid more or less everywhere. … On the other hand, the plot is kept compact and tends, by its triangular shape, towards minimum width.” Furthermore, “As for the rationalization and its corollary - regularity - they must be correctly understood. The triangular contour obviously does not produce inequalities: there will necessarily be acute angles that are very difficult to design, especially for apartments. Finally, whatever one does, the plots will always be different. The idea is not to attain a good English-style uniformity. In many cases one finds large plots in the corners and in the center of the block. One finds plots crossing through and opening on two streets in
the narrowest parts and, beyond widths of approximately 30 m., normal plots opened only on one side towards the street.” (pp. 31-32). One of the best examples of a corner solution is the small restaurant at the intersection of the Rue Rovigo and Rue d’Isly; a remarkable architectural solution to a level change is the Cognon building between the Rue Dumont d’Urville and Rue de Constantine, a composition based on the triangular theme that has a famous precedent in a building by Alessandro Antonelli in Turin, Italy. 84. It is important to emphasize the difference between axis of symmetry and dividing line. An axis of symmetry always presumes movement along it, even when such a circulation pattern is not realized. Thus an axis of symmetry entails a structure with an odd number of bays, such as a basilica plan or a domus plan. In the case of an even number of bays, the central line, which will coincide with the structural bay wall, will be a dividing line, as in some Arab mosques. It is obvious that an axis of symmetry contributes to the organicity of the building. 85. It is interesting to compare the facades on rue d’Isly (now rue Sadaoui M. Seghir) and on the first part of rue de Constantine, with the ones on Boulevard de Sebastopol in Paris. On the new design of the tenement house facades, Hautecoeur writes: “The facades of the tenement houses changed during the second half of the 19th century. Continuing the 18thcentury tradition, architects wanted to give the building a monumental aspect, and used the ground floor and the mezzanine as a sort of base on which to raise two or three floors of decreasing heights in order to build a rhythmic composition made of pillars and crowned by a heavy cornice supporting the balcony of the last floor. The social transformations and economic necessities, however, determined a shift in the organization of the different floors and in the dimensions of the apartments: since the building now had the same social destination at every level, differences in the design of each floor were no longer necessary. The architects were no longer designing according to vertical orders, but according to horizontal elements: the facade finds its rhythm in the balconies, cornices, etc., often continuing in the adjacent buildings.” See L. Hautecoeur, Histoire de l’architecture classique en France, vol. 7 (Paris: n.d.), p. 249.
86. Two cases of highly specialized residential buildings are discussed in: A. Petruccioli, “Polarity and Antipolarity in the formation of the XIXth century City,” in A. Petruccioli, ed., Rethinking the XIXth century City (Cambridge, MA, 1998), pp.8394. 87. C. Aymonino, G. Fabbri and A.Villa. Le città capitali del XX secolo. Parigi e Vienna (Rome, 1975). The most exhaustive text on Paris houses is: C. Daly. “ Maisons de Paris” in Revue Gènèrale de l’architecture et des travaux publiques, Paris, 1859. See also F. Boudon et al., Système de l’architecture urbaine. Le Quartier des Halles à Paris (Paris, 1977). 88. In general, see F. Beguin, Arabisances (Paris, 1983); on Algiers, see X. Malverti, “Alger, Mediterranée, soleil et modernité,” in Architectures Françaises d’Outre-Mer (Liège, 1992).
131
the building tissue Memory is redundant: it repeats signs so that the city can begin to exist. Italo Calvino, Invisible Cities
aggregative principles Illustrations in books and articles are often devoid of any trace of human habitation or the adjacent buildings that form a structure’s surroundings, although in reality a residential building is rarely found that stands isolated from the urban fabric. Many of the examples already mentioned are not individual buildings, but aggregates1 of simpler types: the rab of Cairo, for instance, is a complex of row houses to which access is gained from an elevated exterior path; aside from being above street level, it is not unlike any urban neighbourhood. Similarly, the eighteenth-century apartment house can be read as several overlapping fabrics served by vertical distribution systems (the most evident example of which is the tower). Finally, in the typological development of the domus, the insula process simply changes a type in the building tissue: the courtyard house is reduced to an aggregation of elementary cells. The phases of the mutation of type are interrelated and constitute a continuous process that both affects minimum and maximum scale and is in turn affected by them - thus the need for context. The ambiguity encountered in cataloguing the various examples confirms the non-serial relationship of these successive phases of mutation. We have defined “type” as the organic synthesis of a set of building features in a given time and space. Now we will examine relationships among individual buildings in a broader context. Change of scale involves not only a dimensional expansion but also a simultaneous extension of the interrelationships among building type and building tissue. From the point of view of the level of structural complexity of the tissue, the first intrinsic feature of a building type is its reproducibility. It is a necessary condition, but not a sufficient one, for its qualification as a “type.” A cottage can be replicated an infinite number of times, but the result will be a discontinuous building tissue like a suburb. Thus a type should also include an ability to associate with other buildings of a similar nature so that its aggregation is realized in building tissues and parcelled layouts. Continuous aggregated systems place emphasis on the fundamental phenomenon of architectural and planning iteration. The law obeys local predilections and the historical moment, in a series of open or closed elements, where it enters the conformation of the whole system, qualifies it and singles it out. By iteration I refer not to the standardization processes fostered by industrial production, but to the idea of the coordination of
132
after amnesia Fig. 82 Endogamous building tissue in the village of Tigzirt in Kabilia, Algeria. This neighbourhood, of recent construction, is located towards the outside of the peripheral areas. It is a hybrid between the row-house and courtyard type linearly developed towards the valley. C, original nucleus; B and A are successive duplications with extensions in height and exterior staircase. In A the staircase is outside and allows direct access to and from the public route. F is a “modern” addition (compare with figure 36).
dimensional units as a rhythmic factor that is composed according to the laws of repetition and symmetry and generates a constant repetition of forms in space and time. The rhythm of architectural design today is, unfortunately, little more than a mechanical iteration, a far cry from the aesthetic freedom that rhythm enjoys in music. There are countless examples, from the obsessive syncopated beat of Ravel’s Bolero, or, in Arab music, in which the narrative structure follows iterative patterns, which have neither an overture nor a finale, like the infinite perspectives along the aisles of the Friday Mosque of Cordoba or Omar in Cairo.2 At the higher scale of the neighbourhood, it is meaningless to retrace
the building tissue the matrix type back to the elementary cell, the smallest common denominator. The ultimate module in the retracing process needs to be an urban unit of a larger scale that I identify as a building lot.3 The lot, a minimum urban cell formed by a building and its pertinent area, is the building type tied to its own ground. Within the lot the important concept is the pertinent area, the open space belonging to a given house and destined to absorb the growth of the building. The concept of pertinence or belonging refers not to the principle of ownership, but to the concept of being in a place, settling, dwelling. The concept is spatial in a broad sense as the center of balance pertains intrinsically to a geometric figure. In the area of pertinence the connection is direct; one can enter every point of the pertinence space without leaving it. We can understand the idea of the pertinent area with a simple comparison: in a row house lot aligned along a street, the pertinent area will be the backyard - the open space oriented towards the inside of the urban block; in a Victorian terrace house lot, where the building is usually set slightly back from the street, the pertinent area comprises not only the backyard but the front yard as well, with an occasional, small light well for the basement. In a courtyard-house lot, the pertinent area coincides with the courtyard itself. In the Arab city, the space of pertinence is regulated by the fina, which, depending on the width of the street, can give the front of the house a pertinence from 90 to about 130 cm. The fina is closely tied to the processes of commercialization, where the need to protect merchandise from the elements and the desire to appropriate the public space in front converge, with operations that first follow temporary wood characteristics that are then concretized in plastic-masonry structures4. In a jump in scale the concept of pertinence becomes progressively ambiguous: if the sense of strip pertinence relative to a small building fabric is clear for a row of houses, it is less so in the case of a group of apartments in a block and still less in the case of patio houses. We will see in the case of the city that some tend to identify the area of pertinence with administrative limits. The area of pertinence of the city should be more correctly defined as the group of the grounds of the city, and at the same time, we can say that territory pertains to the settlement/city insofar as it intrinsically contains it. This signifies a revolution of the concept of type, not as a settlement object, but as settlement + ground of pertinence. The aggregation of different lots is the function of the residential type: thus row houses will have a front for access and for a view to the outside and a back facade open to the surrounding space; enclosed houses will be built along the lot boundaries sharing the exterior walls with other houses of the same type, with their facades turned toward the inside. There are exceptions to aggregation mechanisms and the resultant pertinent areas. For instance, in the village of Tigzirt in Kabylia, the neighbourhood unit - a hybrid between the row house and the courtyard type - is composed of two rows of parallel houses of one module only, linked by a sloped path, and closed by the most recent building at its lowest point. It is in reality one house, inhabited by a patriarchal family, the
133
Fig. 83 Routes and blocks built with row houses: (A1) block generated by the progressive construction of a matrix, a planned, and a connecting route; (A2) formation of an encroachment tissue using the corners of the pertinent area; (B) and (C) pertinent strips along a matrix route, and along a planned one with completed encroachment. (From G Caniggia and P. Maffei, Lettura e progetto [Venice, 1979], p. 138).
134
after amnesia
nucleus of which is reproduced as the family grows. In one of the most interesting examples at the entrance of the village, the oldest nucleus C (see figure 82) is laterally duplicated forming a small courtyard; the latter grows uphill towards A until it reaches the public path, then downhill again through a series of duplications to end with the new building F. In this way, the sequence of courtyards resembles an interior street interrupted by walls and doors. Building F appears from the outside as a big box containing all its rooms under a flat roof; the plan is ambiguous, not European: it is subdivided in two by a corridor leading to all the rooms before opening up onto the countryside through a simple connection consisting of an exterior stairway. In reality, it is the covered continuation of the sequence of the courtyards. Thus, the “modern” building is only a reinterpretation of the traditional distributive scheme hiding the coherence of the type behind its out-of-place architectural language.5 All these examples suggest that the elementary cell and the lot are the morphological sub-modules of every building tissue, Obviously, this is
Fig. 84 Ground-floor survey of the Santa Croce quarter in Florence. Note the row-house building tissue composed by narrow and deep Gothic plots. Compare to the courtyard tissues in the following figures. (From L. Benevolo, Storia della città, 2. La città medievale (Bari, 1993), pp. 278-79.
the building tissue not a case of associating the quarter to a Lego-like formal fitting together, but of a building type that is manifested in its aggregate properties and mechanisms in a larger system, like the projection of a sociological unit in which spatial and social values are preserved. The coexistence of different buildings causes reciprocal dependence, as when, in an ancient fabric, the demolition of a house requires the bracing of the adjacent ones.6 The principle of serial aggregation was the basic element in the formation of all Arab fabrics, as Roberto Berardi concluded from his case study of the Tunis medina.7 According to Berardi, the fabric is reducible to a few elements that can be combined: cell, bent vestibule, door, courtyard, and path. The commercial fabric is generated by a simple operation: either the parallel distribution of a series of two cells facing each other along a path (the linear suq), or an envelope distribution around a courtyard (the funduq, the inn.) (fig. 101.3). Such aggregates can meet other aggregates to create various combinations of forms through which the linear organism defines a surface. In this case, the intersecting cell is decomposed into two smaller cells, each belonging to its relative suq. Networked suqs are visible, for example, in Kairouan in the area between Bir Barrouta and el-Bey mosque, and in Fez in the so-called Qaysariya neighbourhood, between the ancient maristan and the Karawiyin mosque, where the el-Attarin suq expands to form a fabric of ten parallel suqs (fig. 101.1). The striking seriality of the Arab city is easily readable in the bazaar, which is not a street as commonly understood, but a sequence of segments, until recently separated by doors: a series plus another series, etc., like the modules of a submarine. This serial pattern and its corollary, the exchangeability of the position of the different elements, are also at the origin of the reversibility of processes: in the medina of Tunis, Berardi has noted that when, in the second half of the eighteenth century, the solidarity among craft guilds diminished and manufacturing declined, the cells of the funduq opened up towards the exterior to become a bazaar. Ideally, the perimeter walls drew closer to one another, progressively reducing the open space to zero: the former courtyard was turned into a common wall between two cells (fig. 101.2). In its general principles, Berardi’s thesis agrees with our reading; it presupposes the seriality of the Arab city and is verified in the simplest urban fabrics of the region like Monastir, Sousse and Sfax, but in Tunis it does not take into account the nodalities that make the picture more complex. He rightly acknowledges the presence of a spatial order eminently structured in an urban framework often defined by others as “disordered”, “spontaneous,” or “arbitrary,” and when he remarks that the plan of a city “is a sort of writing articulated by its morphology and syntax, and evident in its formative process on the condition that interior and exterior, module and aggregation, are all simultaneously considered.”8
135
136
after amnesia
on the urban fabric At first sight nothing seems to resemble Eudoxia less than the design of that carpet, laid out in symmetrical motives whose patterns are repeated along straight and circular lines, interwoven with brilliantly coloured spires, in a repetition that can be followed throughout the whole woof. But if you pause and examine it carefully, you become convinced that each place in the carpet corresponds to a place in the city and all the things contained in the city are included in the design, arranged according to their true relationship, which escapes your eye distracted by the bustle, the throngs, the shoving. Italo Calvino, Invisible Cities On the definition of fabric and on its individuation, urban morphological studies diverge: both Levy9 and Caniggia use the term urban fabric (tissue in French; tessuto in Italian) without defining its parameters in terms of scale, so that for them urban fabric could include both a medieval contrada formed by only part of a street between two rows of houses and a whole urban neighbourhood. In reality, the urban fabric and building tissue rarely coincide. M. R. G. Conzen further articulates the reading scales and the concept of fabric: the plot and the plot series or strips (what Caniggia calls “pertinent strips,” i.e., a serial aggregate along a street), further decomposed into the subcomponents of street system, plot pattern, and building pattern, all coming together to form the superior plan unit. The urban unit that Conzen defines can contain more than one block, provided the morphological components are all homogeneous. The plan unit is a hierarchically superior system that Conzen calls “plan division.10 Individual integuments are plan divisions of the “second order.” The plan division would then correspond to the quarter, provided it shows a relatively homogeneous internal morphology. The crucial problem is how to categorize sufficient levels of scalar complexity to explain all phenomena of aggregation. In this sense, the different steps proposed by Conzen are insufficient, because they do not include all the steps involved in structural complexity. One can define the typical building tissue as the organic sum of common features characteristic of many aggregates in a given unit of time and in a defined area: its formation obeys typological rules and categories, comparable to those characterizing the building type. “The building tissue relates to the aggregate as the building type relates to the building,” says Caniggia. “Building tissue, in this case, can be defined as the conceptual basis of coexistence among buildings that exist a priori in the minds of the designers and builders, a notional concept operating at the level of spontaneous conscience and producing a civic operation putting more and more buildings together”.11 Applying the same principles, one can also postulate the existence
the building tissue of a typological process of building tissues - processes connected to the progressive mutation of the concept of the aggregate over time and in the same place, and dependent on its variants induced by a specific location. However, this does not seem chronologically parallel to the type process. Even if it were, it could proceed at a different speed because hierarchically superior entities have greater resistance to transformation. So while the building type will change quite easily with the passage of time, the building tissue will show a greater resistance to change, the town will display a resistance greater still, and the region the greatest resistance (almost as if to spatial augmentation there corresponds an extension of time).12 It is useful to confirm the concept that the city is founded on a plan with a division into lots that is a function of the common type. In the successive phase there are two fabrics: a new one based on lot division that is a function of the present common type, and a pre-existing one. The present common type will always be out of phase with respect to the fabrics of the preceding phases, and any attempt at adaptation will produce synchronic variants of the type. Every kind of building tissue has its building type preference; in other words, every kind of tissue is formed using the type most appropriate to it. In the most ancient urban fabrics, this principle of simultaneous relationship is valid, from what has been witnessed, only at the moment of their foundation, because almost immediately after that synchronic variations will occur. Even then there can be incongruent cases: for example, when the design of a fabric presupposes a type economically and technically unaffordable by the local society, or when the realization of a tissue begun with a certain building type takes so long to complete that it is finished with another one (the leading type’s synchronic variant of the final period). The first case is exemplified by the reconstruction of the towns of Calabria after the 1783 earthquake: in Polistina, Filadelfia, Seminara, etc., a Neoclassical urban design based on square blocks was realized with leading building types that were in harmony with the local social and economic development, but fell far short of the magniloquence of the original plan.13 The second case offers many examples for the principle that often sees the execution different from the project. The 19th-century expansion of Bari and the coastal cities of Puglia like Trani or Molfetta planned in 1810 during the time of Gioachino Murat were realized over the 19th century with updated building types.
137
the building tissue starting from one pole need in their turn to mediate between their internal structure and the structural necessity of being perpendicular to the route, and will consequently produce a considerable number of trapezoidal lots. In Algiers, Amidouche and Bab el-Jedid Streets, both located on two main ridges of territorial origin that detach themselves from the promontory of Fort l’Empereur, are two matrix routes that have generated the high part of the city (fig. 86.2). The piedmont route in Algiers from Bab Azzoun to Bab el Oued is a matrix route that corresponds to the main suq of the lower city. In the Arab city, the matrix route is often associated with the linear suq. These routes are easily identified on a cadastral map by the minute dimensions of the sub-modules of the stores, serially organized in rows along its sides. This characteristic confers on the linear suq a marked flexibility that allows it to absorb the deformations of the urban fabric. The linear suqs in Tunis are on the east-west tract formed by the Rue de la Kasba, Suq el-Blaghjia, Suq el-Attarine, and Suq el-Trouk, and the northsouth one is constituted by the Suq des Femmes and Suq des Etoffes; in Kairawan, an example is found on the avenue Bourguiba, starting from Bab el-Tounes. In Essaouira rue Ibn Khaldoun is an ancient matrix route of territorial origins, which ended at the inner harbour. Today it is completely commercialized in the route from Bab Doukkala to rue Agadir and reinforced by five important nodes of the Jama Abdellah, Zaquia Amboukiya, Jama Bouakir, Zaquia Jarouliya and the T-shaped node singled out by the Zaquia Nacirya, the Jama al-Agadir and the bath. Each node singles out a passage from the matrix route to the planned route (fig. 136). In Tangiers a matrix route is the segment of the Rue des Chretiens and the Rue Ben Raisoul, connecting Petit Socco Square to Dar el-Makhzen. In the so-called spontaneous town, the routes seldom intersect at 90 degrees because of the odologic principle (see, for example, a plan of Kairawan), but rather tend to form acute angles cutting out triangular lots. In this case the matrix is recognizable among many routes because its sides were built earlier and therefore present regular lots along the whole pertinent strip starting from the crossing, causing the geometric deformation of the strips belonging to the secondary routes. In Kairawan, a similar case inside the city walls (right by the gate of the Boulevard Idris Premier) is composed of two streets, of which the Rue Sidi Abder Moumen is unequivocally a matrix. In Seville,18 a paradigmatic case is
139
Fig. 87 Kairawan, Tunisia: triangular block between rue R. Ben Klaf and rue el Kedidi. The first is a matrix route, the second is a planned one; the third is a connection, due to the typical encroachment shape of the plots. On the intersection of the fork is a later addition of encroachment cells. Also note the presence of a zawiya on the connecting side, and the infill in the inside of the block featuring a primitive version of the courtyard type.
Fig. 88 Examples of planned routes and tissues: (1) block bounded by Rue Rabah Rajah, Rue K. Zenouda, and Rue Ahmed Allem, Algiers; note the deformation of the building types caused by the curving of the contours. (2) The axis starting from Rue Abou Mohajer, Essaouira, Morocco; note the rotation of the small mosque and some early stages of the taberna process on the southern side of the street. The insula process of courtyard houses with a division between ground floor and upper floor determined the formation of an independent stair along the public street.
140
after amnesia Fig. 89 Seville, Spain, in its transition period from a classical to an Islamic city, as visible in the present-day plan. Note the trace of the covered canal, the forks revealing the gates in the city walls (probably from the late imperial period), and the main routes (dotted) converging at the Great Mosque. (1) The cathedral and the Giralda tower; (2) city gate in the Santa Catalina area; (3) city gate in the Plaza Maldonados; (4) bridge on the covered canal in Calle Campana, (5) Alameda de Hercules.
Fig. 90 Palermo. Examples of connecting tissues.
represented by the division of Calle Feria (matrix route) and Calle Malaver starting from Montesion Square, an intersection determined by the presence of a gate in the ancient Roman walls of the adjacent Maldonados Square. Generally, a given pertinent strip corresponds to a symmetrical one on the other side of the route, except in the case of a natural cause of asymmetry such as a river, a channel, or an urban perimeter. However, if for any reason the matrix route is set back from the natural element, the pertinent symmetrical strip will be irregular and segmented towards the rear. This seems to be the case in Seville around the Alameda de Hercules, an ancient pond that dried up during the first Islamic occupation; for obscure reasons its banks had never been built up: the Calle Joachim Costa
141
the building tissue
Fig. 91 Three examples of routes in restructuring tissues: (1) Rue d’Alger in Blida, Algeria, determines a series of trapezoidal plots (at bottom); (2) the construction of the walls east of Kairouan, Tunisia, induces the rotation of the tissue and the opening of a restructuring route cutting the preexisting pattern; (3) detail of the block bounded by the Suq el-Blat, Rue Sidi Ali Azouz, and Rue du Mufti, Tunis. At bottom, note the 45-degree cut of Rue Djamaa Ghorbal, and the bisected mosque.
1
3
2
142
after amnesia
(matrix route) passes tangentially to it, thus cutting out a lens-shaped block - an inverted matrix. In eighteenth-century Cairo, judging from the plan published in the Description de l’Egypte, the Birkat al-Azbakiyya’s sides presented two different patterns: on the north the matrix tissue of the Coptic neighborhood with the front aligned on the waterside, and on all the other sides inverted matrices.19 In Burano in the Venetian lagoon, the central street (a filled-in canal) and the Mandracchio canal have a similar type of route. Another generally applicable rule is that the distance from the pole is inversely proportional to land and building values along the matrix route. Building activity will therefore be concentrated in all available open areas near the pole through secondary routes that will tend to be orthogonal to the matrix: the so-called planned building routes. The distance between two planned building routes varies according to the leading building type adopted, and therefore is specific to the local culture.20 Another distinctive feature of planned building tissues is the lack of facade or access at the beginning of the route, since the two pertinent strips begin at the end of the pertinent strips of the matrix route, and the first part of the planned route is therefore already faced by the sides of buildings of the matrix route (see figure 88). For the same reason, in the presence of the courtyard house where the building lots tend to be square and facade openings are an exception, a reading is more difficult. Some clearly planned building tissues are exemplified by the street network of Damascus, with roads such as Rue el-Djarone, Rue el-Askari, and Rue el-Mahmass cutting the Suq al-Midan every 40 meters to the south; by the modular streets cut on both sides of the Shari’ el-Muizlidin Illah beyond the Bab Zuwaila in Cairo; by the Jdeide neighbourhood beginning with Al Mitran Farhat in Aleppo in the Ottoman period (fig.59); and in Algiers by a block between Said Kaid Soualah, Rue Nfissa, and Rue Smala characterized by a double courtyard system in which the central planned route is later closed to form a cul-de-sac (fig.152), as well as the hill portion of rue Bab el Jedid. Another very clear tissue of planned building routes are the roads with modular passages carved from the principal suq of Fez al-Jedid between Dekaken Gate and Semmarin Gate. The general development of the city and the limits to the expansion of a planned building tissue will lead to the formation of connecting routes. Connecting routes can develop in one of two ways: either through the demolition of contiguous lots on the planned route, recognizable by the fact that the lots will be parallel to the new route, or through design after completion of the block (closure), in which case the lots will be perpendicular to the new route. In the Arab city it is rarer and is recognized only where there are particular constrictions like the presence of walls and hence the need to go from one part of the city to another without converging at the centre. Connecting routes appear late in the Arab city as a consequence of a higher level of urban development, the need for more direct links between urban poles, and the functional specialization of blocks closest to the city centre. Pressed by sudden population growth, plans are made to reduce the dimension of the blocks through the demolition of buildings and the
Fig. 92 The quarter of the Musalla in Damasco. One can note the hierarchies of the routes and the gates of the units of the quarter.
the building tissue
143
opening of new restructuring routes. In the large blocks formed spontaneously, the principle of tree-like penetration almost always prevails, in which the roots tend to expand and in fact avoid connections. The most common tendency is the branching out of a planned route from a matrix route, a possible further Y-shaped branching out of a secondary planned route from the main one, and a further subdivision in a layout of a lesser order, rarely finding connection routes among these. Each hierarchical passage is marked by a gate, and the road section is in proportion to the flux of users and tends to become progressively smaller, down to 90 cm. in the alleys that lead to individual houses. Each hierarchical level of the roads has a specific name: 1. Al-Tarik= Shari’=Shari Al-Azam=suq; 2. Al-Darb= main foundation street leading to a sub-nucleus; 3. Al-Zuqaq= alley or cul-de-sac; 4. Al-Khawkha= vaulted passage 21. In place of the tree metaphor we could use that of the urban aqueduct with the distribution of water into channels of progressively reduced sections and the sluice-gates corresponding to the gates. In the Bab Musalla quarter in Damascus, the network of streets is articulated beginning with a single main street and several derbs, local planned roads that originate in the former, zuqaq or culs-de-sac, and finally some cross streets at the scale of the city. In a short space we find three neighbourhood gates, toward the hill, the valley and the east, five gates that control the passage from the shara to the durubs that innervate the five subnuclei of Sueika, Darzi, Tamina, Nemur and Riashni, all located to the west of the main street, and finally four gates that control the passage to the multi-family house from the durub to the azzika streets.22 In the Chabanat quarter in Essaouira from the matrix suq the planned routes progressively spread out into roads of minor planned routes with a progressive reduction of the road section-down to the fifth hierarchical order. The typological behavior of routes is analyzed according to the recomposition and integration of the elements along them. Cities of medieval origin which have developed linearly through serial duplication and few hierarchies often show a different evolution in their fabric in a clearly readable way: examples are the lagoon fabric of Chioggia and Dubrovnik, and the Cannareggio neighbourhood of Venice. In progression the coordination of the area of pertinence defines the block in successive phases, confirming the hypothesis that, in spontaneous processes, the built route must be anterior to the block. As Kropf pointed out,23 the theories of Conzen and Caniggia are obscure regarding two points, causing uncertainties in the interpreting phase. First, the problem of “shared streets”: what happens to a route when the tissue on one side behaves differently from the other? The reading of a route also varies with the scale: If we consider the aggregative unit or first urban nucleus as an autonomous foundation entity, it is quite probable that the behaviour of the tissues on both sides is analogous, and, vice versa in the area of a very stratified building tissue, it is possible to register a different behaviour with a formal and functional autonomy, as we can see in many medieval cities. This ambiguity cannot be resolved outside a careful contextual reading.
Fig. 93 Essaouira. Block between Derb el Majboud and Derb el Chabanat in the area of Bab Doukala. Note all the lanes blocked by later encroachments ( Thesis project of M. Barbone, A. Brancaccio, S. Mastronicola, M. Minerva, N. Moschetta, M. Stefania, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari)
144
after amnesia
In fact, it cannot be automatically inferred that, along the same route, two pertinent strips facing one another will show symmetrical typological behaviour; on the contrary, one can behave according to the rules of the matrix route, and the other can form a secondary connective tissue. This can be explained with the mechanisms underlying the formation process of the block: if the urban fabric is composed of blocks spontaneously developed over time, the opposite side of the pertinent strip with matrix layout will be completed over a different time, taking the form of a connecting route composition (see figure 90). The tissues of shared streets are important because, if they complicate the reading, inasmuch as they fragment the continuity of those that seem to be long homogeneous routes at the overall scale of the city, they help to configure the chronology of the individual blocks and larger and adjacent parts of the urban fabric. For Kropf, the second ambiguity concerns the formative process of the block: for Caniggia such a process does not exist as a form a-priori, but as the result of a progressive association of building tissues. However, a block in Timgad, city of Algerian limes (fig.121), or in Anjar, Umayyad foundation in Lebanon, is clearly a constitutive element or, in other words, a form of urban design deliberately chosen a-priori. This ambiguity can be solved only by considering the constituent urban block as the arrival point of a process according to which the resultant block at a certain moment is codified as a type of its own, with of course all variants and local differences. Actually the ambiguity does not exist. The rab of Cairo is an example not only of an urban block consolidated in its measurements, but very specialized in its functions, so much so that it often coincides with a single building, similar to the Florentine Renaissance palace. The resultant block is generally the product of spontaneous conscience, while the constituent block is a product of critical conscience; mutatis mutandis, it is the same idea as that behind the formation of the Venetian tissue based on the calle-corte. In the case of fabric in an intermediate stage of development, only an analysis of the formative processes to determine whether all pertinent strips were built simultaneously or not can lead to a precise answer. Urban edges or planned blocks, but growing at a later time, are typically uncertain cases: in these areas blocks rarely form, unless we define as a block the large areas identified by the external radial roads. The restructuring route is characteristic of mature urban forms: it is superimposed over an existing fabric when a new urban pole grows. It is a product of critical conscience, because the financial resources required for such an operation are well beyond those available to an individual or small group. It is for this reason that fabric was rarely restructured in pre-industrial societies, but such projects proliferated in Europe when the modern states began to invest in the city’s image. Traumatic in nature - for it entails a great deal of demolition - a restructuring route is easily readable both on cadastral plans and on the facades of buildings: on the one hand, cutting extant fabric at an angle, it constitutes a new matrix route entirely bordered by triangular and trapezoidal lots of small and medium dimensions and, on the other hand, due to the different period of construction, the building tends to be of
Fig. 94 El-Oued in the Souf region, Algeria. Example of semi-regular courtyard tissue: such regularity is not formally planned, but the result of a collective decision favoured by the flat topography.
Fig. 95 A closer view of El-Oued in the Souf region, Algeria, shows a neighbourhood with courtyard types. Note the strong serial layout of the cells inside every enclosure; each cell enjoys the maximum formal autonomy with its own semicircular or hemispherical covering.
the building tissue
145
146
after amnesia
Fig. 96 Reconstruction hypothesis of the evolution of a domus block in an area where the courtyard type predominates. At bottom is a block of large domus with a 60-foot front, and a maximum depth of 120 feet. Note the phenomenon of commercialization and occupation of the most important public roads. (A) Subdivision of patio houses with the splitting in half of the front and the opening of a path. (B) Substitution with large courtyard houses in the centre of the block, and with smaller houses towards the outside; formation of culs-de-sac. (C) Occlusion of the ancient route.
147
the building tissue a more recent leading type. Special buildings are usually located on the restructuring routes, since a building of high value compensates for the expense of demolition and reconstruction. As special buildings tend to be isolated, on restructured routes they gain the additional privilege of being located all by themselves on lots somewhat smaller than the usual city block. The cadastral plans of the Boulevard de l’Opera in Paris, Boulevard de la Victoire in Algiers, and Rue d’Alger at Blida (fig. 91.1) are clear examples of such a mechanism at work. Although most often associated with Haussmann and the urban design of Paris,24 restructuring routes are common to all times in the history of urbanism, and are always the product of an urban plan. Probably the most unusual ones are in Florence in the via dell’Anguillara and the Borgo de’ Greci, cutting the fabric that developed inside the ancient Roman amphitheatre.24 In the Arab town, always characterized by a more organic development, they were rare until the first half of the eighteenth century, unless we consider as restructuring the frequent rehabilitation of alternative routes in Syria during the Umayyad period, when they took advantage of urban degradation to cut through consolidated Hellenistic tissues. Not even the Ottomans undertook any demolition projects on a scale comparable to Haussmann’s. The State preferred to leave the image of the city to the important people of the empire, for whom it was more convenient to intervene according to architecturally determined poles and nodes that constituted the usual waqf, rather than according to rectilinear cuts. It is not surprising, then, that much of the eventual restructuring occurred around nodes rather than along axes. An exception is the diagonal route in Tripoli between the Arch of Marcus Aurelius and the Bab al-Hurria that takes the name Shara Arba’a Arsat and Shara Jama al-Druj, planned by the Ottomans
Fig. 97 The demolitions and new aligments of the Plan A. Gutton, 1952 at Aleppo ( From Gaube H., Wirth E., Aleppo. Historische und geographische Beiträge zur baulichen Gestaltung, zur sozialen Organisation und zur wirtschaftlichen Dynamik einer vorderasiatischen Fernhandelsmetropole. Wiesbaden 1984, fig.11).
148
after amnesia Fig. 98 Tripoli, cadastral plan with architectural survey of the special buildings. One can see the demolition of the Turkish era corresponding to the Shara Arba’a Arsat and Shara Jama al-Druj. a. Arch of Marcus Aurelius; b. Crossroads of Arba’a Arsat; k. Gurgi house; j. Karamanli house; l. Mohsen house; o. Bait al-Mal; 1. Jama al-Naqah (courtesy of Ludovico Micara)
149
the building tissue in 1551 after the reconquest of Tripoli along with the reconstruction of the nearby Jama al-Naqah mosque. Studying the tissue west of the Shara al-Drug it is evident how the Ottoman demolition profoundly destroyed the existing urban fabric. We also note the creation of a parallel street, the Shara Sidi Hamura, which shows the intention not only to create a new connection, but to form a new planned fabric as well.25 Major urban transformations often had enough room left over for what have later become real restructuring axes. André Raymond tells us how in Cairo the moving of the dyeing factories, which formerly occupied an area of four hectares southwest of Bab Zuwayla, favoured the global intervention of the Ridwan Bey waqf between 1629 and 1647: the project included a sumptuous palace, a tenement building, a caravanserai, two zawiyas and one sabil, and redesigned the street for 150 meters including the famed covered bazaar, the Qasaba Ridwan,26 and the northern building tissues. In Aleppo three major transformations of the Hellenistic fabric south of the bazaar, coordinated between 1544 and 1583 by the Ottoman governors, led not only to the addition of several hectares of commercial
Fig. 99 Some examples of increments to an original geometrical nucleus of special or residential buildings are: (1) the block bounded by the Rue Tourbet el-Bey, Rue du Persan, and Rue des Juges in the northwest quadrant of Tunis; the central part shows traces of a semi regular cluster settlement, served by a single cul-de-sac. Note how the entrance is also marked by the typical forked layout in front, just like a city gate. Also note the more irregular development tissue that borders houses between the first nucleus and the roads tangential to it. (2) A funduq in the suq in Fes el-Bali. A single increase with a row of shops. (3) Kairawan, Tunisia: building tissue developed according to a curve in the Rue des Tapis;
2
3
1
150
after amnesia
surface, but to the reorganization of the bazaar streets as well.27 In Essaouira tradition has it that the successive extension to the first plan of the city willed by Sultan Sidi Mohammed Ben Abdallah is the work of the French engineer Theodore Cornut, who arranged the cardo-decumanus plan of the two wide commercial streets, the Suq-al-Jadid, extending between the Mashwar and Bab Dukkala, and rue Mohammed al-Qorry, intersecting at 90 degrees in the central market. A careful examination of the behaviour of the tissues over the two routes reveals some surprises: where we would expect a tissue on the matrix route we find the presence of trapezoidal lots, indicators of reconstruction cutting. Given this, it is clear how Cornut’s grandiloquent plan is a more modest attempt to give geometric order to the new city, cutting across previous settlements of tribal origin.28 The situation changed radically in the eighteenth century. André Raymond refers to two cases in Cairo: a west-east cut through the old town to link Muski to the al-Azhar region (today it is called Sikkat el-
1
2
3
4
Fig. 100 Examples of encroachment tissues: (1) the bu lding at the southeast extremity of the medina is an example of infill between two diverging rows of houses, Hammamet, Tunisia. (2) The area to the north of the Ali Bitchin mosque in Algiers, an example of encroachment inside a Roman theatre. 3) The mellah in Essaouira, Morocco: the encroachment of the open areas inside the block has few synchronic variants because of the extreme density of the tissue. (4) Example of an occupation of the central area of a block of special buildings, Kairawan, Tunisia; a zawiyia has encroached on the centre of the block between avenue Bourguiba and rue Homet el Bey.
152
after amnesia
the Boulevard de la Victoire (fig. 151) and the Rue des Consuls, finally reducing it to a fragment within the colonial tissue.30 Tlemcen and Miliana in Algeria had even worse luck: they were cut in all directions by so many restructuring routes that the original tissue was reduced to tiny remnants.31Italian colonial urbanism, in contrast, preferred to avoid contacts with the indigenous cities, a policy which favoured the conservation of ancient tissues, such those of Tripoli and Benghazi and oases like Gadhames and Ghat.32 We should emphasize how the route of modern reconstruction is the death knell of the tissues with courtyards. The jump in scale of the tall buildings that are built on the edges of the new route is an unendurable challenge to the privacy of the courtyards below. In Aleppo, the demolition of tissue like rue Moutenebbi carried out according to the plan of Andre Gouton in 1954 has greatly compromised the life of the adjacent traditional quarters, robbing them of air, light and privacy. We can sum up by stating that the typological behaviour of the building fabrics is tied to the hierarchy of the routes: a careful analysis of the former allows us to define the order and role of the latter in the city. We have identified matrix, plan, connection and reconstruction orders of routes. This last one is a separate category, since it is hierarchically equivalent to a matrix. In the Arab-Mediterranean city, the story is more complicated: it should be noted how, on the one hand, more serial fabrics prevail, whose routes subdivide into minor planned routes, taking the tree form, which can have up to five or six minor branches. It should also be noted how, on the other hand, because of the difficulty of interpreting the courtyard fabrics along the routes, and an ambivalence interpreting their use, the latter often present a hierarchical ambivalence.
153
the building tissue the courtyard tissue In the case of the row house tissue, the desirability of facing the street led to the exploitation of the length of the route, and only later to a progressively denser occupation of the interior of the block. The angle of the latter represents a strong nodality causing the encroachment of the first lots according to a typical step shape (see fig. 83). By and large the extension of the courtyard tissue does not impinge on public space, since it generally prevents building right on the street. A first analysis identifies the courtyard tissue as being of relatively low density, especially in rural areas or in those historical centres little affected by population growth. From aerial photographs of the large traditional cities like Fez, the courtyard tissue appears as a sequence of voids that balance the solids.
1
2
To optimize their orientation, courtyard aggregates tend to be laid out along east-west routes so their open space will face south. The matrix routes and planned routes, whether or not they are laid out north-south, behave simply as general traffic paths; rarely are lots distributed along them. Rather, they are distributed with their longer dimension or with their building fronts parallel to the street. It is the secondary planned routes that distribute lots according to the optimal orientation: parcels will therefore be laid out parallel.Variants to it are generally caused by specific topographical conditions: in the case of a change in ground level, in fact, the series of courtyards have to steer between the need to maintain their linearity and the need to adhere to the curvilinear path of the topographic line. They tend to keep the routes to the latter parallel, while climbing paths (or sometimes stairways) will not be parallel, but slipped diagonally in a gradual way forming a fan-shaped or serpentine fabric. In the higher part of the Algiers Casbah, where there are traces of a parcelled layout based on the Roman actus, we can read ancient walls that are rotating while always remaining perpendicular to the contours, and consequently aligned with off-axis stairways located at every ten-meter level change. When the rotation of the contours could not be absorbed by the perimeter walls, the triangular left-over lots were occupied by encroachment tissues (see fig. 150).
Fig. 102 Two examples of a fork: (1) the path north of Rue Sidi Abder Moumen is clearly a matrix route, Kairawan, Tunisia; the planned routes are legible from the deformation of their first plots. (2) The intersection between Rue Saida Ajoula and Rue du Divan, Tunis. Originally the fork featured a widening before the corner which has turned perpendicular with the infill of a special building. Rue Saida Ajoulae is still readable as a matrix route.
154
after amnesia
a
b
c
d
All courtyard tissues can be reduced to three basic models: 1.Tissues with open series: linear monodirectional tissues that could theoretically grow infinitely. These aggregates have remained at the lower stage of development typical of rural areas, where to every row of lots there is a corresponding path. The courtyard type varies according to orientation and access; open series tissues behave asymmetrically: on the north side there is only one long wall with the openings leading to the court, and to the south the dwellings present a blind facade, especially at the beginning of the block (see figure 105). 2.Tissues with a double layout or closed series: usually in proto-urban settlements, they consist of a route for every two rows of contiguous lots (fig.94), to give access to the courtyards either from the unbuilt side (therefore with central access) or from the building (in this case with a lateral access leading to the court through a covered passage). They are particularly common in Mediterranean Arab towns that have more or less regular layouts and in linear or crossed forms with alternated parallel routes. Because the second is the more frequent case, we can define the others as synchronic variants of it. In addition, there are other important examples in Fez el-Jedida, in which the rhythm is always constant (20-22 meters) despite its orientation.33 Both linear and regular double layout tissues can be found in aggregates of more recent origin, such as those outside Bab Souika or Bab Djedid in Tunis; outside Bab el-Hadid in Aleppo; in the northern neighbourhoods of Bab el-Tounes, and in the new blocks next to Jami el-Kebir in Kairawan (see figure 140). 3.Deep or profound tissues. A solution in particular peripheral or topographical conditions when it cannot be served by a second parallel
Fig. 103 Diagram of aggregations of the courtyard house. (A) Open series tissues: linear unidirectional tissues are able to develop indefinitely. Typical of rural areas, they are the aggregates which have remained in the most undeveloped state. Each row of plots is served by a path. (B) Double-layer tissues: consist of a path between two rows of contiguous plots, so that courtyards are accessed either from the unbuilt side -- thus with central entrance -- or from the building with a lateral entrance leading to the court through a covered passage. (C) Common variant of the first model: a peculiar solution in those peripheral areas or topographic conditions that cannot be served by a successive parallel route. This solution leads to the formation of a dead-end. Despite the orthogonal layout of the routes, the courtyards maintain the orientation of the preceding ones. (D) Simultaneous presence of the former cases.
Fig. 104 (right page, above) Village in the Haouz in Marrakesh, Morocco. Example of spontaneous courtyard tissue formed by aggregation along pre-existing rural routes. Fig. 105 (right page, below) Aerial view of a ksar in the Ziz river valley, Morocco. Regular building tissues and streets almost all covered by sabat.
the building tissue
155
156
after amnesia
route is found when it is desirable to use an area behind an already existing one: this is achieved by designing some culs-de-sac provided with secondary paths. Despite the orthogonal angle thus formed, the rear courtyards maintain the orientation of the preceding ones. Not even the lots at the intersections escape this rule. Only occasionally do such paths depart from the north-south route and, even when they do, they are not followed by the construction of residential buildings on that side. In the central areas of the city in the large blocks the first type of fabric is arranged on the edges and the third occupies the interior. Instead, the closed series fabric is typical of villages beyond the first construction where the formation of forks prevails. Although developed spontaneously, these tissues are rather regular, the result not of an authoritarian planning strategy, but of a general rationality in building and the agreement between proprietors having equal property rights. The subsequent phase in the development of the tissue is the process of filling in open space between the first building tissue and the main routes: this generates a belt of houses surrounding the original settlement. Thus, the rule that the most ancient layers will always be towards the centre of the block is still valid. The periphery is open public space and functionally more flexible and more subject to frequent change than the centre. The block bounded by Rue Tourbet el-Bey, Rue du Persan, and Rue des Juges in the northwestern quadrant of Tunis perfectly shows this process of cluster settlement: semi-regular, of homogeneous social class, and developed all around one cul-de-sac only. The area also shows the subsequent phase of urban-route modification-those that have bent pass tangentially to the settlement’s corners-and the belt tissue developed between the first settlement and the routes themselves (see fig. 99.4).
Fig. 106 Block bounded by Rue du Mufti, Souk elBlat, and Rue Sidi Ali Azouz in Tunis; note the Hinna funduq, a perfect example of encroachment on the central area of a block composed of special buildings.
158
after amnesia
rules dictated by the gromatici veteres. Cities of Hellenistic origin like Aleppo, Damascus, Homs, Latakia, Antakia, Alexandria, Edessa or Roman ones like Jerusalem (destroyed and re-planned by Emperor Hadrian),35 Jericho,Tripoli, and Fatimid Cairo, Algiers, and many others show evidence of an original urban or rural fabric based on classical dimensions. In these cases the problem becomes one of studying the typological process of spontaneous courtyard tissues superimposed over a planned tissue based on modular courts. Jean Sauvaget first reconstructed the development of Latakia36 and of other important Syrian cities,37 and summarized it into a schema showing the progressive encroachment of colonnaded monumental avenues (that could be called super-matrix routes) leading to their transformation into a suq, but he did not decipher the capillary transformation of residential tissues in the rest of the city. After him, for more than fifty years in studies of the Islamic city no one was able to reach the successful synthesis between archaeological data and an interpreting of the planimetric forms. A more recent essay by D. Sack38 applies historical data to a descriptive analysis to delve deeper into the rules underlying aggregates. However, this can be done today thanks to analytical comparisons between many more examples than the few available sixty years ago. Syracuse will be used as the paradigmatic example; the other case studies will be described only through analytical drawings. The Fatimids took Syracuse39 after a long siege that ended with the death of its entire male population, and occupied it for more than two hundred years (878-1086). We can therefore infer that they appropriated a classical fabric and that it underwent the typical transformation processes -courtyards in the process of being divided up among many families, encroachment onto the grand porticoed avenue, and a series of incremental or shrinking phases that still maintained the original boundaries as an overall constant. All over the Ortigia island in the old city centre, a very dense fabric was laid out over a strigae grid in which dwellings, if compared to archaeological plans existing nearby (Imera, Megara Iblea, and Selinunte), confirmed the hypothesis that they were pastas houses - a square enclosure 15 meters wide and 26 meters deep - with the elementary cell at the northern edge. The Graziella neighbourhood seems to have the richest potential for analysis: located at the edge of the ancient city, it has a coarser grain suggesting that at the time of its conquest its fabric consisted of a series of peripheral orchards and houses. The circulation system appears as a re-planning of the domus lots of the Roman epoch. The phases of these processes seem to have followed the typological rules already described and still readable today: in fact the neighbourhood (one of the poorest, where the Moors were confined after the Norman conquest) did not undergo any substantial development until the fifties of this century. This neighbourhood was characterized by houses of one or two stories at most, aggregates of elementary cells not developed in depth.40 It is easy to imagine how every Islamic clan, having gained possession of one or more domus already in the process of becoming insula, put them together to form a unitary neighbourhood filling the voids; they extended a path thanks to the demolition of a wall or a single cell. This path would then serve the residential areas deep within the site of the original domus.
Fig. 108 The Graziella quarter in Syracuse. From top to bottom: traces of the structure of the Roman foundation from 21 BCE; elementary domus with Greek plan; processes of commercialization and encroachment; processes of the use of the courtyards (from M. Zampilli. Lo sviluppo processuale dell’edilizia di base, in A. Giuffré, op. cit. pag. 45-46).
the building tissue Flexibility of the pre-existing fabric such as that found in Syracuse explains the extraordinary linearity with which a city in the Mediterranean area was able to form and develop over another without a rupture. We can imagine and partly reconstruct the analogous mechanisms that allowed continuity of settlement in all Mediterranean towns, with the general rule of thumb that the quantity of information we can assemble from their reading is inversely proportional to the layering and specialization of the fabric. The more layers to the fabric, the more the interpreting needs to be integrated with archaeological and toponomastic evidence.41 Finally, we can identify another type of very common urban aggregate: the encroachment tissue. This is a synchronic variant of the main building tissue composed of smaller or simpler versions of the leading type, in a way that at times resembles shantytowns in developing countries today. It usually developed under difficult topographical conditions, such as an infill development area at the bottom of a valley; in harsh topographies after the easiest and most advantageous locations had already been taken; at the center of large lots built in different phases and by social classes other than the ones inhabiting the perimeter buildings carried out through the progressive subdivision of the lots; and finally, frequently in the form of encroachment on public space. The progressive transformation of the domus into the insula can also be considered a sui generis encroachment tissue.
159
Fig. 109 (Above) Meknès, Morocco. A view of a zuqaq in the Touta neighbourhood. Below) Meknès, Morocco. Zenkat Sidi Abdallah el-Kesri, connecting route in the Touta neighborhood, view of the street with a sabat and cantilevered constructions.
160
after amnesia
A peculiar form of encroachment occurs in corners and is determined by the nodal value of the corner position, and the consequent densification by subdivision to which its pertinent area is subjected. This phenomenon usually involves more units, determining the typical step shape of the corner lots, and is most common in those building tissues comprised of row houses (see fig. 83). Obviously, there is nothing morally wrong with encroachment tissues. They represent a spontaneous social action of adjustment to a pre-existing planned tissue. A plan is the product of average values, and it goes without saying that it cannot represent the totality of needs and desires of every individual in a community. The encroachment, an occupation of precious public property or of less desirable areas within the urban context, represents the instinctive reaction of the spontaneous conscience and a desperate pragmatism.41 In Tunis, the medina sector bounded by the Suq el-Blat, the Rue du Mufti, and the Rue Sidi Ali Azouz has been studied by Berardi42 (fig. 106); it perfectly shows a case of encroachment composed of special buildings. This building block represents a coexistence of residential and productive activities around a wealthy-family nucleus dedicated to textile manufacturing since the seventeenth century. The centre of the block,
Fig. 110 Cadastral plan of the city of Aleppo, scale 1:500, section 5, sheet 2 area of Bab Qinnesrin.
the building tissue probably once private open space, was entered through a door on the Rue an-Nayyar; it had been gradually encroached upon by the Funduk de la Hinna. It is still home to textile manufacturers employed by the Ghorbal family. The houses inhabited by the family are all interconnected, turning the funduq into a self-contained neighbourhood. The same Rue Djemaa Ghorbal clearly originates from two culs-de-sac connected by a diagonal path that also intersects the homonymous Ghorbal mosque. In the fifteenth century, Kairawan was redesigned when the great zawiyas were added. These buildings for religious brotherhoods were housed in architectural complexes composed of numerous courtyards around which were gathered the patron saint’s tomb inside a qubba, a madrasa, a prayer hall, and one or several houses for the family. Because they needed large open spaces to accommodate pilgrims, the zawiyas were usually located at the edge of the city or, like the Zawiya Sidi el-Alwani, at the center of large urban blocks, encroaching little by little on all the open space43 (fig. 100.4). Since encroachments are the last formations over time, in the analysis they must be the first to be distinguished and eliminated to make the picture clearer.
161
Fig. 111 Hammamet, Tunisia. Example of the reading of a block in the medina: (A) isolation of the plots; (B) isolation of the types and synchronic variants with relative measurements; (C) entrances and original bayt (all houses are oriented in the same direction); (D) layout of services; (E) territorial density of each building unit. (By Youssef Chennaoui, student project in a specialization course, “Analyse typomorphologique,” taught by Attilio Petruccioli, Ecole polytechnique d’architecture et urbanisme, Algiers, 1991-92).
162
after amnesia
interpreting building tissue and urban fabric The most useful documents for interpreting both building tissue and urban fabric are aerial photographs, cadastral surveys, and architectural measured surveys or forma urbis. The first two should be at a scale sufficiently large to show individual urban lots - a scale no larger than 1:2000 (photogrammetric maps at 1:1000 or 1:500 are the most common today). This information can be combined with projects realigning routes that were designed by technical public bureaus in the 19th century. The cadastral maps, in those versions indicating all parcels, are particularly useful for analyzing the urban implant and its variation over time. They were made to classify landed property (land cadaster) and building property (urban cadaster) and were introduced in Europe in the second half of the 17th century.44 They recorded every property with a plan survey (usually in 1:100 scale or equivalent), and apartments or units were inventoried on separate sheets to be assembled later on a general city plan at 1:1000 or 1:500 scale.45 A cadastral plan shows the number of divisions, the totality of surfaces, the layout of open spaces, and alignment of roads. On it one can perform geometrical operations and introduce physical and spatial relationships between the elements represented. Since cadastral maps act like a snapshot, they can be used to verify the modifications that collect in a fabric over time. On a cadastral plan one can isolate special buildings,
Fig. 112 Reading the bu lding tissues, the Baroque trident in Rome: The tissues on the matrix routes are in black; the planned tissues are in a grid pattern; the connecting tissues are dotted, and in a finer dotted pattern the restructuring tissues, as Via della Fontanella di Borghese. The disposal of the matrix tissue underlines the sequence of growth starting from Via Ripetta, and then Via del Corso and last Via del Babuino (see also the scheme in figure 115).
the building tissue
163
enclosures of holy buildings, aggregations within a tissue, and so on. But above all the land register is the land property materialized on the ground, represented by closed surfaces, generally the private space called lots. Geometric criteria permit the distinction to be made between the lots that tend to be rectangular and then trapezoidal, as deformations of the rectangle because of the presence of contour lines in the land or restructuring routes, or triangular. The L- or T-shaped lots derive instead from the process of subdivision and acquisition. The French colonial administration was especially interested in cadastral cartography, and one of the commonly acknowledged masterpieces of the French administration are the cadastral maps of Syria published in the 1930’s, and in particular the one of Aleppo and the one of Antakya, from 1931 and 1929 respectively.46 The cadastral maps of cities made in the second half of the eighteenth
Fig. 113 The Chabanat quarter in Essaouira with its hierarchy of routes appears as a closed point served by a tangential circular street, from which the culs de sac begin that converge in the centre (Thesis project of A. Cutrone, P. Genchi, Gm. Lozupone, L. Massarelli, A. Moccia, D. Panaro, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari).
the building tissue
165 Fig. 115 The medina in Essaouira, Morocco: reading the building tissues with the addition of cul-desac tissues (dark hatching). The distribution of the restructuring tissues created by the plan of Theodore Cornut, a cardo-decomaneous scheme which cuts through pre-existing tissue. It is therefore not a plan with new foundations, but a rationalization of existing urban nuclei. In solid black, matrix tissues; in crossed lines; foundation tissues; in dotted lines, connecting tissues.
shown in fig. 84, which illustrates the underlying serial layout of long Gothic parcels) is definitely a process that will yield unclear results when applied to an Islamic fabric made up of lots, in which one of the dimensions is not much longer than the other; the pertinent strip will be even less readable in those cases in which the block is composed of clusters or single buildings extending in depth only. In other words, the medieval Florentine one is serial, and the other partially organic. To take apart a tissue, there are two complementary operations of interpreting, one an analysis of the functional and formal relationships between the plot series and route, and the other a dismantling of the elements of the neighbourhood tissue. In this phase we are not considering special buildings since their super-module has often overwhelmed the fine grain of the residential fabric, making the picture illegible. In the first operation, it is better to proceed according to successive levels of typological specificity, recognizing first the pertinent strips and their relation to the route. This frequently ends in the discovery that the depth of the pertinent strips is generally constant along the whole route, thus clearly distinguishing the encroachment tissue behind. The depth of lots varies according to the individual expressions of each building culture (and thus of the type adopted): in Rome it is usually between 24 and 30 meters, but it can reach as much as 42 meters along some main routes; in Essaouira, it never exceeds 15 meters; in Aleppo, with some exceptions, the two dimensions are almost equal - that is, the lots are as deep as they are wide, as a result of the adopted courtyard type. In Essaouira, the 15-meter measure seems almost a standard in all its planned building routes and in the clusters that build up the central blocks; the width is reduced to 10 meters along the restructuring routes like Suq al-Jadid, and finally to a
166
after amnesia
narrow 3-6 meter strip along the matrix routes, as in the tract bordered by the Rue des Zair and the Rue Ibn Khaldoun. The next step is to catalogue the typical features of the lots along the various categories of route - planned route, connecting route, restructuring route, and cul-de-sac - to obtain at the end a mosaic of the fabric in which all building tissues of the urban sector under analysis are represented. Such a mosaic is an invaluable document, for it permits us to read the various behaviours of tissues and their respective synchronic variants: it allows us to reconstruct a synthetic image of their typological processes as an essential premise to the subsequent reading of the urban evolution via the hierarchical relationship of its nodes and poles (see figures 112 and 115). Suppose that three roads converge at an obligatory passage - a bridge, say, or the gate to a city: at their vertex they will form a trident cutting out quadrangular and triangular blocks in the urban morphology. Now consider only the two extreme cases that (1) the three routes are contemporary and the product of a precise plan; or (2) the three routes were traced at different times. In the first case, in the hypothesis that the central axis links the passageway to an important pole (as represented in fig. 112), building tissues would behave as matrix routes on both sides; the A and C axes would behave as planned tissues on their inner sides and, in the absence of other contextual conditions, as the matrix of future tissues on their exterior sides. Closer to the vertex, where the triangular lots are smaller, the connecting routes either will not form at all or will tend to merge with the planned routes starting from the central axis. It is as if the matrix route were reduced to a point - the convergence of the fork or trident- able by itself to generate a planned route. The phenomenon is present in the rabads or outlying suburbs like the long block of the R’bat bab Souika in Tunis, or the block between derb el Majboud and derb Bab Doukkala at Essaouira that tend to acquire the form of very narrow triangular blocks (fig. 93). In the second case of a successive formation - for instance beginning with A and ending with C in figure 112 - matrix tissues would be clearly defined on both sides of the A axis, perpendicular to which planned routes and their relative tissues depart at regular intervals; blocks end along the inner side of the B axis with connecting tissues. The sides of the axis would exhibit a different behaviour: on one side, there would be a diagonal connecting tissue dictated by the orientation of axis A; on the other, a matrix tissue generating other blocks in the direction of axis C, which then follows an analogous process. The variation in the chronological order of the formation of the three axes will produce a different mosaic of tissues, which we can now logically decipher in their chronological and morphological evolution. Figures 112 and 115 show two mosaics of the fabric from very different contexts of the six typical building tissues: the first is the quarter of the Tridente Barocco in Rome, developed in the small parcels of the urban design made by the monks of San Silvestro in Capite;49 the second is the urban settlement of Essaouira built west of Marrakech by the sultan Sidi Muhammad b. Abdallah in 1764,50 and the third is Jerusalem within the walls.51 The first two were chosen for the relative simplicity of the layout
Fig. 116 Diagrams of the formation of building tissues in a trident. At top, in the hypothesis that B is a matrix route, the whole system develops symmetrically towards the side streets. At the bottom, in the hypothesis that A is the original matrix route (in this case, the Via della Lungara in Rome), the system grows progressively from A to B (Via del Corso) to C (Via del Babuino).
167
the building tissue
2
1 and for the moderation of their changes over time. The clear reading of the first case, with the exception of an ambiguity in tissue where the Via Condotti intersects with the Via del Corso, is less evident in the second case. In Essaouira different morphologies coexist: from the regular grid of the Old Qasbah and the diplomatic quarter south of Rue Allal ben Abdullah, to a semi-regular fabric without any apparent geometric layout such as the quarter around Rue Khald ben el-Qualit behind the Place de France, or by the quarter of Beni Antar and Rue Agadir, until we reach the most organic areas corresponding to the first nomadic settlements, like the Bawakhur and Chabanat quarters. This city does not present many connecting routes, with the exception of the ring around the perimeter composed of the larger blocks toward the northern walls. A problem in Essaouira, and in the Islamic world in general, lies in evaluating building tissues generated by culs-de-sac. Should they be considered as a specific tissue having their own formation rules, as anthropological studies would confirm? Or should we consider them as either a planned tissue, or as a secondary connecting tissue? Is it not contradictory to consider a cul-de-sac as a connection at all? In effect they are planned routes sui generis since they are drawn beginning with a principal route or another planned route; they penetrate into a fabric, but fold, forming a broken line because of the “resistance” they encounter along the way in the owned lots.
Fig. 117 Examples of culs-de-sac: (1) Western side of Via S. Agostino, Palermo; four lanes define the depth of the block. (2) In the mellah, a small neighbourhood of five houses, Essaouira, Morocco.
168
after amnesia Fig. 118 Derb in the Mokhfiya neighborhood, Fez, Morocco; a route in the form of a fish skeleton leads to two main houses; small lanes branch off to serve the smaller houses of the clients.
The mosaic thus obtained is the necessary premise for a diachronic reconstruction of the transformations of the hierarchy of the routes and the relative tissues. Once the eventual persistence of agricultural land is recognized and isolated, we move on to the recognition of the bands of routes parallel to the direction of development, like the territorial routes and successive perimeters of the walls. In the second operation the various types of tissues are isolated in relation to the various routes. The tissues are also characterized by phenomena such as incremental growth, fractioning, occlusion of routes, and encroachment which have to be isolated as well. At the level of the building, in every tissue all the building types that contribute to its formation are separated. The elementary type can rarely be isolated in evolved tissues as an independent unit of dwelling, but only as a module of more complex entities. The various units are studied in relation to the size, depth, and access of the tissue. As for the block, we can notice that a depth longer than the typical one is common in peripheral areas, exemplified in Rome by large special building types like monasteries and in the Arab city the cemetery areas and their relative pilgrimage points. It is also important to mention that the synchronic variants of some blocks - specifically those that present a reduction of their depth in order to adapt to specific cases, like pre-existing structures in their pertinent strips, or specific topographical conditions - determine an analogous behaviour of the building tissue, therefore of the type.
the building tissue
169
By the end of this interpreting phase we should have classified the different building types and building tissues, and have synthetically represented their reciprocal relationships in a non-hierarchical form. In all likelihood our method is not able to give back all the layers of information offered by the survey, since it must break down the built reality into its discrete elements and the simple aggregations of these, which we can think of as modules used by the builders of the city over time. The reading will not be destructive of the overall unity of the object if it keeps in mind that each module in itself is not interesting if it is not placed in a universe, which is the plan of the city.
Fig. 119 Kairawan, Tunisia: three derb in the medina. The uniformity and low density of the upperright-hand derb on the southern city limits indicates its recent construction.
170
after amnesia
cul-de-sac The cul-de-sac played a crucial role in the development of the cities in the Mediterranean area influenced by Islamic culture. The cul-de-sac is a path of varying width (from a minimum of one meter to a maximum of three) without an exit, linking a public street to a residential neighbourhood that belongs to the third model of courtyard tissue. It has been given different names in different Islamic countries, some of the more common of which are zuqaq, zanqa, derb ghayr nafidh, and sikka gahyr nafidha.52 It is formed by two parallel walls, usually blind or with a few small openings, a sort of open-air corridor leading to several private entrances with doors that never face each other so as to avoid visual inspection. Walls are sheer, anonymous surfaces, the entrances (the only point of contact with the outside) are profusely decorated. In the past a door often separated the zuqaq from the rest of the city. Today, although such diaphragms have long disappeared, the inhabitants still make access by outsiders very difficult.53 The morphology of the zuqaq can be very complex: straight sections can suddenly take a right-angled or even a U-turn; covered parts alternate with open-air ones, bordered by houses of differing heights, fostering different feelings of physical compression; occasionally it functions as the neighbourhood core, often housing a common service like a fountain or a well; at other junctures houses bridge the zuqaq, turning it into a tunnel. The infinite gradations of light of the streets of Gadhames in Libya guide the visitor and become ordering elements of the hierarchy of spaces. The light allowed into the streets through a system of wells from above is directly proportional to their section and importance: we thus go from small piazzas, where reverberation during the day prohibits all associative life, to the street of the bazaar measured by the light that floods in from above, with its rhythm of light pouring in, to the total darkness of the culsde-sac that lead to the houses, where only the owners are able to move with ease like moles.54 Enrico Guidoni sees in the cul-de-sac the tissue he considers the primary formative element for the urban morphology of Islam; he regards it as an intentional design having a strong iconic significance and extrapolates a formal catalogue of Arabic elements for it from the Sicilian towns of Arab origin.55 Such a deterministic model seems unlikely; the cul-de-sac is always integrated with the formation of deep courtyard tissue. We should not exclude the possibility that it was a device deliberately chosen to provide access to the centre of the block between the houses aligned along the public route when they were built, or even to provide for the future development of an interior cluster distant from the matrix route. Other deterministic interpretations of the cul-de-sac as an exclusive product of a specific meteorological condition are equally unconvincing, although it is unquestionable that a narrow street is more shaded and better protected from desert winds than a wide one. Neither does the association with military defence seem satisfactory, although safety is admittedly a central theme in the layout of medieval settlements in the Mediterranean area. It seems more reasonable to study the origin of the cul-de-sac as a mechanism for exclusion connected with the desire to contain inter-ethnic
the building tissue
171
Fig. 120 Aleppo, Syria: complex neighbourhood unit outside Bal al-Hadid. The neighbourhood lowest in the drawing had been specialized with the building of a mosque whose doors open onto opposite culs de sac. (Redrawn from A. Gangler, Ein traditionelles Wohnviertel im Nordosten der Altstadt von Aleppo in Nordsyrien (Berlin, 1993), table 13).
conflicts and preserve a degree of tribal autonomy within the city core. In the Islamic town, this fundamental attitude could have been jealously conserved by the social unity of the group inhabiting the cul-de-sac until well into the twentieth century. In the past, the conservation of the status quo was guaranteed by using Divine Law to watch over social behaviour: from the juridical point of view, the cul-de-sac was a private domain governed by the rules of Islamic justice summarized by the fundamental principle, “Love thy neighbour”. Several scholars, including Brunschvig56 and Von Grunebaum,57 have attempted to derive principles of urban and architectural composition from Islamic law, with results that were often not worth the effort. The five schools - Malaki, Hanafi, Shafi ’i, Hanbali, and Jaffari - of Islamic Law diverge on nodal points: the selling or the alienation of air and light (the degenerative principles underlying encroachment tissues) are not
172
after amnesia
prohibited by the Malaki school of North Africa, for example, but are strictly forbidden by the Hanafi of Turkey. From the Mahkana shar’iyya of Cairo, which provides numerous applications of the Law of the Koran, L. Fernandes58 extracted the following principles that directly influence urban form: (1) an entrance door cannot be shared by two neighbours; (2) one entrance door cannot face another; (3) mashrabiyyas or sabat are not allowed without the approval of all the residents surrounding the cul-de-sac. To these we can add the convention of limiting the minimum zuqaq width to four cubits, the equivalent of a loaded camel, and the minimum height to seven, equaling a camel with rider. These very simple rules, comparable to an elementary building regulation, are integrated with more general principles59 that directly influence building, such as those regulating the drainage of water and polluting substances, and the universal right to privacy. It is evident that the legal codes condition social behaviour and preserve ancient customs, thus contributing to the evolution of building types.Yet, to assume that the religious-juridical laws dictated the form of the tissue (and specifically culs-de-sac) directly and that a syntax or a grammar could be derived from them is far from credible.60 Moreover, with the notions of space in the two Mediterranean Islamic cultures that are exact opposites - the introverted Arab one originating in the oasis and the extroverted Turkish one formed in the high plains - we can infer that the law adapted to the idea of already existing space, and not vice versa. In other words, since the concept of space is a deep cultural acquisition measurable only in longterm perspective, it is logical to suppose that the various ethnic groups were already building their settlements in the way described before their conversion to Islam. We can sum up by stating that the Sharija codified uses and customs tied to the courtyard house through its coherence with the sense of territory; the khitta, the tribal law of the distribution of localizations, adopted it for its climatic and social efficiency; the crisis of the Dar al Islam after the 14th century, closing off the tissues, exasperated its growing mechanisms, multiplying closures and culs-de-sac, but the courtyard house and the cul-de-sac are not products of Arab tribal law, but a technique for organizing autonomous space. The cul-de-sac and the cluster are aggregate types intimately linked to the courtyard-house type developed in the Neolithic period. Examples abound, from the Ur quarters at the end of the third millennium BCE, to those of Assur and Megiddo around 1750 BCE, and of Deyr el-Medina in Egypt. We can find them in Ahmedabad, India, in the Padana plain in Italy, and Campidano in Sardinia, with differences due to local typological processes. Even where it is possible to follow the evolution of classical tissue, like the ones from the medieval to the present-day Maghreb, the cul-de-sac clearly appears as the result of a process of adjustment to an existing scheme. On the contrary, the correct interpretation is intrinsic to the building type and its aggregation in the building tissue. The cul-desac is the most reasonable solution to serve houses in the interior of deep urban blocks. The large urban blocks typical of an early settlement shorten the length of the routes, if the access is organized with a cul-de-sac. It is
the building tissue not by accident that the subdivision of the large urban blocks determined by the progressive stratification of the fabric occurs in the Arab city at the expense of the cul-de-sac with an increment in the number of routes. To sum up, we have here highlighted the social character of the building type and its purpose in being aggregated with similar types. It is true that some familiar building types like the cottage and the hut resist aggregation, but in those cases we are not dealing with an urban form but at most with a “primitive village.” American suburbs in this sense are not different from primitive villages. Introducing the notion of movement in the evolution process of the tissues not only allows us to understand the sequence of the formation of the routes within the urban fabric from the matrix route to the planned route to the connecting route and finally to the restructuring route, but it also permits us to logically connect the two processes of movement and tissue. Encroachment tissues, even if forgotten by scholars of urban history, are important, visible in the difficult topographical places where synchronic variants of the leading typical tissue operate. The false impression of disorder should not be turned into a moralistic judgment: the tissue of encroachment is a form like any other and usually represents an individual reaction to a rigid plan. Furthermore, through a large number of recorded cases we have learned to recognize changes in the building tissues either in relation to the formation of routes or to internal changes. Finally, the problem of the formation of the cul-de-sac has been discussed, demonstrating that even if it is a mechanism found throughout the Islamic world, it does not pertain exclusively to it.
173
174
after amnesia
notes 1. “Aggregate” is the general term used by Caniggia to indicate an ensemble of buildings (G. Caniggia, G. L. Maffei, Composizione ... la lettura, p. 122). This concept may be thought of as the highest common denominator in arithmetic. The unitary module is no longer the smallest physical unit of the tissue; rather, a composite of the smallest unit along with some of the surrounding environment forms the new base unit, so that the whole tissue may be resolved into multiples of the new unit. 2. In classical music, repetition can be a theme in and of itself, as in compositions known as “ostinato.” Arabic music is serial, and like the decorative patterns in architecture, it involves rhythm and melody but not harmony, and it does not allow itself to be set like western music into closed systems, with a prelude and a finale. The musical piece can in fact be interrupted at any time during its execution. Arabic music is based on modal homophony and on the darb rhythm; from the 13th to the 16th century the modal system consisted of twelve principle modes, or maqamats, each mode being a cycle of notes to be repeated ad libitum. The rhythmic modes, from the original four in the Umayyad epoch, increased to not fewer than twenty-one in the 16th century. H. G. Farmer, History of Arabian Music (London, 1929), pp. 71-72. 3. According to Conzen, a plot is a parcel of land representing a land-use unit defined by boundaries on the ground; see M. R. G. Conzen, Alnwick, Northumberland: A Study in Town Plan Analysis (London, 1969), p. 128. Thus Conzen associates plot with use, whereas Caniggia implies it. We should integrate this general definition with the concept of pertinent area (see G. Caniggia, G. L. Maffei, Composizione... la lettura, pp. 122, 129-130). Plot and pertinent area are not always immediately readable in the urban tissue: on the contrary, we can say that their relationship becomes more ambiguous in 18th-century tissues and more blurred in Modern Movement plans. In Le Corbusier’s Plan Voisin, every X-shaped tower stands on one plot, which equals one block; similarly, the slab-block building occupies a whole plot, corresponding to one block, or two blocks when it bridges a street.
4. The concept of fina was developed in R. Brunschvig, “Urbanisme Médiéval et Droit Musulman,” in Revue des études islamiques, 15, 1947, pp. 127-155 and Salih al-Hathloul’s book The Arab- Muslim City (Riyadh, 1996), pp. 94-102, with references to main legal texts, among them Ibn al-Rami ,the Muhtasib of Tunis with the best illustrative examples. Fina can be defined as semi-private collectively owned spaces alongside individual properties considered to belong or at least to be susceptible for use by residents of these properties. 5. See B. Weuters, Jr., Construire un autre village (Leuven, 1986). 6. G. Cataldi, ed., “Edilizia seriale pianificata in Italia 1500-1600,” Studi e Documenti di Architettura 14 (Florence, 1987). 7. R. Berardi, “Lecture d’une ville: la Medina de Tunis,” Architecture d’Aujourd’hui 153 (December 1970-January 1971), pp. 38-43. 8. R. Berardi, “Signification du plan ancien de la ville arabe,” La ville arabe dans l’Islam (Paris, 1982), p. 187. ; R. Berardi, “L’architettura delle città nelle epoche di formazione dell’Occidente Cristiano e dell’Islam, “ in A. N. Eslami, ed., Architetture e città del Mediterraneo tra Oriente e Occidente (Genoa, 2002). 9. Levy defines it as a system composed of many elements (reseau): street network, land division, buildings, open space, context, and the subcomponents of topography, hydrography, and vegetation. The syntactical relationships of the reseau define a certain urban tissue, within which it is possible to distinguish the single components, the urban facts. In our opinion, this concept of tissue is too general. See A. Levy, “Forme urbaine, tissu urbain et espace public,” Morphologie urbaine et parcellaire, ed. Pierre Merlin (Paris, 1988), pp. 93-98. See also idem, “Contribution au projet urbain. Introduction à une théorie de la composition urbaine,” in V. Spigai and A. Levy, eds., Le plan et l’architecture de la ville (Venice, 1989). 10. Plan unit is defined as “Any part of a town representing a combination of streets, plots, and buildings distinct from its neighbours, unique in its circumstances, and endowed with a measure of morphological unit and/or homogeneity. Plan units represent essentially morphogenetic plan
types and vary in character so as to produce a morphogenetic typology in which the simple combinations represent subtypes, their integration to more complex units forming “types.” Conzen. Alnwick, Northumberland, p.128. The following is Cozens’s definition of a plan division: “A geographical group of morphogenetic plan units, a morphogenetic plan ‘region’ within the town. Urban plan regions arrange themselves in a hierarchy of two or more orders depending on the site and complexity of the town. The kernel or Old Town, with or without its inner-fringe belt, depending on the plan character of that belt, forms a plan division of the first order as does the totality of the integuments outside.” Conzen, Alnwick, Northumberland, p. 116. 11. G. Caniggia and G.L. Maffei, Composizione.....La lettura. op. cit., p. 123. 12. Karl Kropf says about the permanence of landed property: “The strength of the social and legal limits set by the division of land as property is considerable. One of the most striking examples is the effect of the initial subdivision of land on a new settlement or an addition to an existing settlement. … The building gives force to the abstract division of the land. Once built, the physical definition of the plot reinforces the limit set by the property boundary. Being relatively inert, it is likely to remain as built until it is considered of less value than a different or altered form. Karl Kropf, “The Definition of Built Form in Urban Morphology,” Ph.D. diss., University of Birmingham, 1993, p. 139. The Italian term lotto derives from the archaic French word lot, which means heritage. Chastel and Boudon demonstrate that urban tissues formed by small lots are greater than tissues formed by large ones (F. Boudon, A. Chastel, H. Couzy, F. Haiou, Système de l’architecture urbaine. Le quartier des Halles à Paris, 2 vols. [Paris, 1977]). Geographers have studied the phenomenon according to which, in the evolution of an urban tissue, functions (land and building use) change more rapidly than the buildings themselves, while the parceled layout and the route network last for centuries. An eternal permanence seems to characterize the structural traces of the territory (unless deliberately erased), as proved by the geometric subdivision of the Roman and Hellenistic land in Syria, Tunisia, or in the Padana plain in Italy; another similar example is the dimensional structure of
175
the classical city, still evident today in the majority of Mediterranean cities. 13. P. Maretto, “Edificazioni tardosettecentesche nella Calabria meridionale,” in Studi e Documenti di Architettura 5 (Florence, 1975). 14. W. Isard, Location and Space Economy (New York, 1956), p. 350. 15. See K. Levin, “Der Richtungsbegriff in der Psychologie. Die spezielle und allgemeine hodologische Raum,” Psycologischeforschung 19 (1934): 286. Modern highways suspended on piles and detached from the ground, connected to their surroundings at only a few points, are the revenge of Euclidian space on the hodologic. 16. M. Poete, La città antica. Introduzione all’urbanistica (Turin, 1958), p. 39. 17. G. Caniggia and G. L. Maffei, Composizione … la lettura, p. 132. 18. For the analysis of Seville, see the photographic survey published in Atlante di Siviglia (Venice, 1992). 19. On Azbakiyya, see D. Behrens-Abouseif, Azbakiyya and Its Environs from Azbak to Ismail, 1476-1879 (Cairo, 1985). 20. Caniggia discovered that in medieval row-house tissues the distance is generally equivalent to twice the depth of the pertinent strip (G. Caniggia and G. L. Maffei, Composizione... la lettura, p. 134). 21. Sources confirm this progressive hierarchical form, without shedding too much light: Ibn Dukmak at Fustat classifies the streets into 1. al-aziqqa: Shari Suk, Shari fi Altarik, Zuqaq Ghaigh Nafiz, Tarik and 2. derb: al’Darb al-Kabir, Dar Ghaigh Nafiz which, with no topographical reconstruction, remain rather obscure; P. Casanova, from the interpretation of Makrizi (“Essai de recostitution topographique de la ville d’al-Foustat ou Misr,” in Mémoirs IFAO, 35, (Cairo, 1919), classifies the streets into shari= main street in general, shari ala al-tarik= artery; tarik= route; darb= street, no different from zuqaq and sikka, and finally khawkha=vaulted passage not different from sabat. 22. R. Thoumin, “Deux quartiers de Damas, le quartier chrétien de Bab Musalla et le quartier kurde,” in Bulletin d’Etudes
Orientales, (Cairo, I, 1931) pp. 99-135. 23. K. Kropf, The Definition of Built Form, p. 198. 24. The amphitheatre is visible in the plan of Florence drawn by F. Magnelli and C. Zocchi in 1783 and published in Firenze nelle vedute e piante (Florence, 1926). 25. Ludovico Micara suspects that the section is the deviation of an ancient Roman cardo, resulting from the destruction carried out along the moat of the Castle. See L. Micara, “Tripoli, Madinat alQadima: un tessuto urbano mediterraneo,” in Tripoli città fortificata del mediterraneo, in A. Marino and S. Ciranna, eds., Ricerche di Storia dell’Arte, 86, 2005, p. 50. A series of indications confirms his intuition: proceeding from north to south we note that the Funduq Zummit opposite the Arch of Marcus Aurelius has a rhomboidal plan that tends to support both directions of Hara al-Kebira, the decumanus maximus and the Ottoman demolition of Shara Arba’a Arsat; at the crossroads of Arba’a Arsat (four columns) the Gurgi house has the wast ad dar oriented with the Roman city, while the eastern side has progressively deformed bayts; the Caramanli house has a double deformation, and in particular the eastern side tends to recover the old Roman alignment, creating a protrusion in the crossing. The rectangle of the block that contains it to the east of the new Ottoman cardo turns the smaller sides in the same direction, distorting itself into a rhombus; the Bait al-Mal house, in the restructured block facing the new direction, features the access alley that is a remnant of the old Roman plan; the small Masjid Ibn Tabib has a rhomboidal plan. The twisting part of the Suq alTurk east of the Jama al-Drug, on the other hand, substantially maintains an alignment parallel to the old Roman cardo (see fig. 98). 26. A. Raymond, Le Caire (Paris, 1993), p. 221. 27. A. Raymond, Grandes villes arabes à l’époque ottomane (Paris, 1985), p. 224. 28. According to the French consul Louis Chenier, the plan was drawn up by Theodore Cornut in 1767. See L. Chenier, Recherches historiques sur les Maures (Paris, 1787). The Cornut plan is kept in the Bibliothèque Nationale de Paris, plans et cartes, portefeuille 110, division 4, rouleau 1012. See also S. Mouline, Repères de la mémoire ( Rabat, 1997); Ben Driss Ottmani,
Une cité sous les alizés. Mogador dès origins à 1039 (Rabat, 2002); D. J. Schroeter, Merchants of Essaouira, (Cambridge, 1988). 29. A. Raymond, Le Caire, p. 302. 30. The most accurate account of sixteenthcentury Algiers is given by Brother Diego de Haedo, Topographia, e historia general de Argel … (Valladolid, 1612), translated into French by Monnereau and A. Berbrugger, Revue Africaine 14 (1870) and 15 (1871). See also D. Brahimi, Opinions et regards des européens sur le Maghreb aux XVIIème et XVIIIème siècles (Algiers, 1978). On colonial Algiers, see R. Lespès, Alger. Etude de géographie et histoire urbaine (Paris, 1930); X. Malverti and A. Picard, “Dalla citta indigena alla citta europea: il servizio del genio e la ristrutturazione degli insediamenti in Algeria (1830-1870),” in Storia Urbana. 35-36 (April 1986): 3-40; A. Petruccioli, Alger, p. 104. 31. On Tlemcen, see R. I. Lawless, G. H. Blake, Tlemcen. Continuity and Change in an Algerian Islamic Town (London, 1976). For a complete documentation on colonial French urbanism, see J. Royer ed., L’urbanisme aux colonies et dans les pays tropicaux, Communications et rapports du Congrès International de l’Urbanisme aux Colonies, 2 vols. (1932-1935). 32. In Tripoli, capital of the colony, both the 1914 and 1931 master plans left the indigenous town intact; they chose instead to lay out the colonial settlement east of the castle along the main routes converging at the Souk el-Kobra (the bread market). The exception was the Arch of Marcus Aurelius at the intersection of Shara Hara Kebira and Zanqat al-Fransis, a work of the architect Florestano di Fausto. For further information, see Giuseppe Miano, “Florestano di Fausto from Rhodes to Libya,” in The Presence of Italy in the Architecture of the Islamic Mediterranean, ed. A. Petruccioli, Environmental Design n. 9-10 (1990), pp. 56-71. Benghazi was also surrounded by boulevards, but they were installed without demolishing anything. On the 1914 master plan of Tripoli, see L. Luiggi, “Porti, spiagge e fari della Libia,” in Giornale del Genio Civile, 1913, pp. 14-235; on the 1931 master plan, see G. Reitani, “Politica Territoriale e urbanistica in Tripolitania. 1920-1940,” in Storia Urbana 8 (May-August 1979): 49-64. Italian colonial urbanism was the subject of an exhibition whose catalogue is particularly
176
after amnesia
informative; see G. Gresleri, P. G. Massaretti and S. Zagnoni, eds., Architettura italiana d’Oltremare 1870-1940 (Venice, 1993). 33. E. Wirth, “Stadtplanung und Stadtgestaltung im islamischen Maghreb. Fes Djedid als Ville Royale der Merinider (1276 N. Chr.),” in Madrider Mitteilugen, 32 (1991): 226. Wirth suggests that even the domus tissue of the Roman city of Volubilis near Fez presents the same modulation. In the Fez mellah especially in the area of Derb al Fuqi, where there are fewer subdivisions of the lots, we find the same measurements of 20-22 meters. See S. G. Miller, A. Petruccioli and M. Bertagnin, “Inscribing Minority Space in the Islamic City: The Jewish Quarter of Fez (1438-1912)”, JSAH, Journal of the Society of Architectural Historians, 3,( 2001) pp. 313- 327. 34. K. A. C. Creswell, Muslim Architecture of Egypt, 2 vols. (New York, 1978), vol. l, p. 126. 35. J. B. Segal, Edessa the Blessed City, Oxford University Press, 1970, and for Jerusalem: Captain Ch. W. Wilson, Ordnance Survey of Jerusalem, 1865, reprint (Jerusalem, 1980); M. Ben-Dov, Historical Atlas of Jerusalem ( London, 2002); D. Bahat and T. Rubenstein, The Illustrated Atlas of Jerusalem (Jerusalem, 1990). 36. Jean Sauvaget, “Le plan de Laodicéesur-Mer,” Bulletin d’Etudes Orientales, Institut Français de Damas, (1934), pp. 81-114. 37. Jean Sauvaget, Alep. Essai sur le développement d’une grande ville syrienne dès origines au milieu du XIXe siècle (Paris, 1941). On Damascus, see idem, “Le plan antique de Damas,” Syria 26 (1949), pp. 314-358. Sauvaget noted that in the eastern part of Damascus to the north of the Via Recta there are two roads with a north-south alignment 400 meters long with a linear and parallel course and two others going in the opposite east-west direction that cross the former at right angles. To the south of the Via Recta are two other roads with a north-south direction that have the same distance between them as the first two. The connection points of the four roads on the Via are strictly equidistant from one another. Only the eastern quarters of Damascus have this web today, but using a grid with the measurements found all over the city, we find many coincidences, short segments of a coherent system that reveal a grid extending regularly over the whole city. The latter has two renowned precedents: K.. Wülzinger and C. Watzinger,
Damaskus, Die antike Stadt (Berlin, 1921); idem, Damaskus. Die islamische Stadt (Berlin, 1924). Sauvaget’s ideas and methodology are discussed in N. Elisseeff, “Damas à la lumière des théories de Jean Sauvaget,” in A. H. Hourani and S. M. Stern, eds., The Islamic City (Oxford, 1970). 38. D. Sack, Damaskus. Entwicklung und Struktur einer orientalischen-islamischen Stadt (Mainz, 1989). In particular, she examines the three urban quarters of Harat Bab Tuma, Harat Bain as-Surain, and Harat Qanawat-Taidil. 39. See R. Martin, P. Pelagatti, G.Vallet, and G.Voza, eds., Storia della Sicilia, vol. 1 (Naples, 1979), pp. 685 ff. 40. See R. Bollati and S. Bollati, Siracusa: genesi di una città: il tessuto urbano di Ortigia, (Reggio Calabria, 1999); F. Cordano, Antiche fondazioni greche (Palermo, 1986); P. Pelagatti, “Elementi dell’abitato di Ortigia nel VII e VIII secolo,” in Cronache di Archeologia 17 (1978); S. L. Agnello, “Osservazioni sul primo impianto urbanistico di Siracusa,” Cronache di Archeologia 17 (1978). 41. An example of an impeccable historical reconstruction based on written sources is S. Denoix, Décrire Le Caire. Fustat-Misr d’après Ibn Duqmaq et Maqrizi (Cairo, 1992). Denoix reconstructs the whole street network, but not according to a geometric layout. The exemplary historical method, based on a comparison of historical sources and urban iconography, is described in detail in “History and Urban Form: A Methodological Approach,” Morphogenesis, ed. A. Petruccioli, Environmental Design 1-2 (1993), pp. 70-81. This text can be combined with “The Piedmont of Blida,” a typological reading by Mohamed Saidi of Mitidja, Algeria, and published in the same issue (pp. 50-59). For the Roman school of ancient topography that employs a similar method, see P. Sommella, “Finalità e metodi della lettura storica in centri a continuità di vita,” in Archeologia Medievali,VI (1979), pp. 105-128. 42. R. Berardi, ‘“Signification du plan ancien”. A careful study of the processes of obstruction of the crusaders’ citadel after the Ayyubide reconquest up to the present day can be found in: M. Bouteflika, TartousSyrie. Lecture stratégique et restauration urbaine, doctoral dissertation in Architectural Design for Mediterranean Countries, A. Petruccioli, advisor, Bari Polytechnic, Italy, (2004). The very accurate survey reveals not
only the blocking of open spaces around the crusader buildings, but the small illegal additions of the upper levels, determined by the declassification of the building fabric from super-specialized or nodal to residential or basic fabric. A capillary study of the so-called new mellah of Essaouira was carried out by M. Barbone, A. Brancaccio, S. Mastronicola, M. Minerva, N. Moschetta and M. Stefania, Progetto di riqualificazione del tessuto storico di Essaouira, Marocco, degree thesis, Politecnico di Bari (2005), A. Petruccioli, advisor. The survey of each building unit has revealed the processes of blockage with subdivision of the type up to the codification of much more elementary basic types. 43. P. Jervis, “Kairouan,” Urban Fabric, special issue of Environmental Design, 1-2 (1989), pp. 36-53. 44. From ancient Greek kata = through and stichos = line, meaning to register. The cadastral survey was systematically developed in the eighteenth century, when land and building became a bourgeois item, thus taxable property. It is not necessarily an objective tool; on the contrary, it is an instrument of power representing reality through deliberate abstractions, a specific code one has to know in order to make use of it. However, since the objective of the cadastral map is to describe taxable properties, it requires precise measurements: thus, the cadastre provides an accurate record of all urban dimensions (see R. Zangheri, Catasti e storia della proprietà terrriera (Turin, 1980). 45. The 1:1000 scale resolution allows the representation of a barren parcel and its boundaries, the street network, and some pertinent areas such as courtyards. In the last century, the French tradition adopted the 1:500 scale: the cadastral plans of Algiers and Constantine (also in Algeria) represent covered parts like sabat and driba, loggias and porticoes. 46. The 1:500 cadastral maps of Aleppo represent the parcel boundaries with a reference number leading to cadastral registers. Its graphic representation distinguishes built areas from open space, the walls shared by two properties, streets with levels, vegetation inside courtyards, presence of water, jumps in level, steps, stairways, iwans, and wells. Roofs are distinguished as flat terraced or double pitched (J. C. David, “La formation du tissu de la ville arabo-islamique, apport
177
notes
de l’étude des plans cadastraux d’Alep,” in Urban Morphogenesis, pp. 138-155). 47. See F. Boudon et al., Système de l’architecture urbaine, op.cit., pp. 75-93 48. The architectural measured survey of ground floors is an Italian tradition that has its precedent in the famous plan of Paris by Vasserot (1820-1840). The survey of the Renaissance quarters of Rome is published in S. Muratori, R. Bollati, S. Bollati and G. Marinucci, Studi per una operante storia urbana di Roma [Rome, 1963]); also important is the survey of Venice published in Studi per una operante storia urbana di Venezia by Saverio Muratori, cited earlier. As for Islamic countries, there are surveys of Tunis at 1:200 scale by the Atelier de la Sauvegarde de la Medina; a survey of Kairouan by Paolo Donati and Paola Jervis; one for Hammamet by Mario Face; and for Sfax by Michel Van der Meerschen, published in Monumentum 8 (1972), pp. 10-11. This author is completing the survey in 1:500 scale of Essaouira. In a research project between MIT and the Orientalisches Seminar of Tübingen, the forma urbis of the sharistan of Bukhara has also been completed. See H. Gaube, A. Gangler and A. Petruccioli, Bukhara:The Eastern Dome of Islam (Stuttgart, 2004). At the Architecture School of Bari in the thesis final studios under my direction since 2000, consistent pieces of the tissues of Mameluk Cairo, Tartous in Syria, Aleppo and Antakya have come to light. The architectural survey can be elaborated with on-site measurements represented at 1:200 scale, or through a patient assembly of all single properties, and plans catalogued at local technical offices, after on-site verification of their measurements. In the Islamic countries it is possible to find urban cadastral plans of the medina divided into several plates, but without plans of the individual properties. In this case there is no alternative but direct on-site survey. 49. The 1:200 ground-floor survey of the Tridente Barocco buildings used for the mosaic is published in P.Vaccaro and M. Ameri, Progetto e realtà nell’edilizia romana dal XVI al XIX secolo, 2 vols. (Cortona, 1984). 50. There are few bibliographical references and no urban history for Essaouira. M. Delaborde and A. Mano, Essaouira. Le temps d’une ville (Casablanca, 1991), contains a brief historical introduction. For other
information, see also D. J. Schroeter, Merchants of Essaouira. Urban Society and Imperialism in Southwestern Morocco 18641886 (New York, 1988). 51. J. L. Abu-Lughod, Cairo. 1001 Years of the City Victorious (Princeton, 1971). The mosaic was assembled by the author according to the cadastral map updated at 1:1000 scale, although incomplete. Reconstructions of the historical topography of Cairo in the Fatimid epoch have been attempted by K.A.C. Creswell, The Muslim Architecture of Egypt,vol. 2: Ikhshids and Fatimids (Oxford, 1952), and by P. Ravaisse, as reproduced in M. Clerget, Le Caire. Etude de géographie urbaine et d’histoire économique, 2 vols. (Cairo, 1934), figs. 28 and 29. 52. Aziqqa existed in Palermo until the 14th century; documents and public registers show quarters and streets of medieval Palermo with Arab shera (shari) and sucac (zuqaq), maintaining this name until the 14th century, reports Vincenzo Di Giovanni in La topografia antica di Palermo dal secolo X al secolo XIV, 2 vols. (Palermo, 1889-1900), pp. 286-287. 53. Eugen Wirth studied the complex system of gates in Fez el-Jedid, showing how the city is in fact composed of adjacent self-contained autonomous sections. See E. Wirth, “Stadtplanung und Stadtgestaltung im islamischen Maghreb,” pp. 213-231. The ideas on the Islamic city of this renowned German geographer are synthesized in one of his most popular texts: “Die orientalische Stadt. Ein Überblick aufgrund jungerer Forschungen zur materiellen Kultur,” in Saeculum 26. no. 1 (1975). See also, by the same author, Die orientalische Stadt in islamischen Vorderasien und Nordafrika, 2 vols. (Mainz am Rhein, 2000). 54. A. Petruccioli, “Ghadames, la perla del deserto,” in Al Farabi, I, 1983. See also T. Bernet, En Tripolitanie:Voyage à Ghadames (Paris, 1912), and L. Eldblom, Structure foncière, organization et structure sociale, une étude comparative sur la vie socio-économique dans les trois oasis libyennes de Ghat, Mourzouk et particulierment de Ghadames (Lund, 1968). 55. A. Casamento, P. Di Francesca, E. Guidoni, and A. Milazzo, Vicoli e cortili. Tradizione islamica e urbanistica popolare in Sicilia (Palermo, 1982). H. Bresc, in his article “Filologia urbana: Palermo dai Normanni agli Aragonesi,” in Incontri
Meridionali, III, 1-2 (1981), pp. 9-40, arguing with Guidoni, emphasizes the more universal value of the alley still present in the Arab quarters of the Albergheria, of the Seralcadi and of the Kalsa in Palermo. 56. R. Brunschvig, “Urbanisme médiéval et droit musulman,” Revue des études islamiques (1947), pp. 127-155. 57. G. von Grunebaum, The Structure of the Muslim Town in Islam. Essay on the Culture and Growth of a Cultural Tradition (London, 1961). 58. L. Fernandes, “Habitat et prescription légales,” in L’Habitat Traditionnel, pp. 419426. 59. General principles that have influenced Islamic urban form are: (1) it is forbidden to invade the privacy of others, including looking onto their properties; this affects the construction of tall buildings, the location of entrances, and especially separation devices like the driba and skifa of Maghreb. (2) Neighbours and relatives have the right to buy the adjacent property. (3) Vertical extension is admitted, although it can block air and light from neighbours, on condition that it does not invade their privacy. (4) The fina, or exterior pertinent area, belongs to the house, but does not entail ius aedificandi; see J. Schacht, An Introduction to Islamic Law (Oxford 1964). 60. This method has been accurately used by Besim Hakim in a historical-urbanistic analysis of Tunis: B. S. Hakim, The Structure of the Muslim Town in Islam. Essay in the Nature and Growth of a Cultural Tradition (London, 1961).
178
after amnesia
Aleppo and its Citadel.
the urban organism
179
the urban organism It is difficult to find any enviroment that is not higly coherent, initial impressions notwithstanding... Such enviromental coherence is ubiquitous, and it is not fully attributable to the presence of hierarchical structures of form and place. John Habraken, The Structure of the Ordinary The third and most complex cycle on the urban-scale hierarchy is the urban organism.1 The term “urban organism” includes settlements ranging from the village, to the proto-urban agglomerate, to the town and the city. Insofar as possible, however, the terms town and city will be avoided, for they reflect ideas that are both culture-bound and in a state of flux. If it were simply a matter of the relative number of inhabitants per territorial unit, the definition of city would be straightforward, but generally the word has not only quantitative but qualitative connotations as well. Many settlements in Italy’s agricultural south, in particular the Apulia region and the Sicilian hinterland, are dormitory settlements - that is, predominantly residential and lacking urban facilities. For example, Andria, northwest of Bari, has a population of 80,000 inhabitants, who work mainly in the fields and return to the settlement in the evening. Its facilities are limited and better suited to a town of 2,000 than to one of 80,000. On the other hand, Como, an industrial center in northern Italy that specializes in silk production, has approximately the same population but has a sophisticated range of facilities that include an opera house, a school for training seaplane pilots, and four dealers in rare books, reflecting its wealthy citizenry. In contrast to Andria, Como’s organism developed into a city, a distinct urban entity, as the social body awoke and expressed its complex
Fig. 121 One domus in block no. 90 occupies half the block. (Below left) Timgad. The plan of the city. (From A.E.J. Morris, History of Urban Form [London, 1974], p. 55). (Below right) Plan of the southwest area of Timgad. (From E. Boeswillwald, A. Ballu, and R. Cagnat, Timgad, une cité africaine sous l’empire romaine [Paris, 1905], table 166). Timgad and Pompei are among the clearest examples of a continuous and interactive relationship among building type, bu lding tissue, and urban fabric. Note the discrete flexibility in the distribution: the corner house of the squared block no.103 occupies one fourth of the block’s surface, whereas a domus with one taberna takes up the entire block no. 91.
180
after amnesia
relationships in physical terms. As a result, it has a representative array of specialized, meaningful nodes in the form of institutions and monuments. Analogously, a large rural settlement of the Fertile Crescent can not be compared to a caravan route city in Syria like Aleppo. These two examples clearly show how complex the problem of defining a city can be. This dilemma has filled the pages of the classic Arabic geographers. Mukadasi in the Ahsan postulates three main types of city: misr=metropolis, kasaba=fortified center and medina=small provincial city, without arriving at a definition of the nature of the city, or the differences among the three types.2 So far we have avoided this complexity by assuming the building tissues to be serial aggregations deprived of the overarching organicity of an urban fabric. No city exists, however, whose parts are not hierarichized: a hierarchy of parts and specialization of elements are the essential conditions for its existence. In the urban-scale hierarchy, structural complexity ranges from the urban nucleus to the urban fabric to urban street/block patterns to the urban organism.
Fig. 122 Aerial view of the Moulay Ismail hill, Morocco, with extremely dense fabric laid out parallel to contours on a hillside.
the urban organism
181
the urban nucleus type As we have seen, the neighbourhood is a module of building tissue that varies in form, depending on time and place. A derb or cluster served by a blind alley in the Arab-Mediterranean city (and the entire pre-Islamic east), a contrada in an Italian medieval city, the calle-corte in Venice and the Dalmatian cities, and a square in London are all kinds of neighbourhoods. Although every neighbourhood is individual and specific, one can move from the single specific building tissue to the typical building tissue by concentrating on the similarities. The neighbourhood no longer acts as the point of reference for a physical and social microcosm, but instead as a module capable of blanketing an urban surface. The typical building tissue, when repeatable and aggregated with other typical building tissues to form an urban fabric, can be defined as an urban nucleus type or urban unity. Urban nuclei exist in an urban fabric in the same way that materials make up an architectural structure and building types form a building tissue. In the Arab-Mediterranean city, with the exception of military elites and notables, the population is distributed in these distinct nuclei of rather homogeneous size, around a thousand inhabitants (hara) in sub-groups linked either by religious solidarity and for Muslims of the same ethnic origin (Berbers, Arabs, Turks, etc.) or according to the four juridical schools, or because they emigrated from the same city, or they are part of the clientele of a noble family, or because they belong to the same tax bracket. Lacking a more complex urban political structure, these are examples of integrated social life. Building tissue made up of a single type is most conducive to the formation of urban fabric. Both tissue and fabric will then be composed
Fig. 123 (Above) Morphology of the medina of Meknès, Morocco. Serial aggregation of typologically similar units moving upward in scale from the courtyard house to the derb, to the neighborhood (haouma), to the urban fabric composed by the urban nucleus (qsma) to the medina. (Below) Analysis of the compositional and functional elements of the Touta quarter, Meknès. In the case of a small derb of equal houses the nodality is concentrated at the gate of the soqaq with minimal facilities. In the case of derb, hierarchized by a dynastic house, the nodality tends to move along to the core of the cul-de-sac. At the level of the haouma (neighborhood) there is a central nodality with public buildings, an oratory, and local public commercial facilities. Higher up on the hierarchy, the qasma involves more facilities like a public fountain and a mosque.
182
after amnesia
of nuclei (neighbourhoods, quarters, streets) that are based on the same leading type or its synchronic variants. This uniformity of composition gives the fabric identity as well as the ability to be identified with the complex and stratified morphology of the city as a whole. On the other hand, modern peripheries owe their monotony and reduced determination to the excess individuality of the single buildings, whose structure answers only to an internal logic that goes against the built context. Pre-colonial Arab and Persian cities provide some of the clearest examples of monotypical composition. They are characterized by an almost wholesale adoption of clusters of courtyards extending up to the city walls. Aerial photographs of these cities show an endless expanse of regular openings, and it is quite difficult at first to spot discontinuities in the fabric. An impression of uniformity is created that is correct from the vantage point of the airplane, but incorrect when one’s feet are on the ground, because neighbourhoods in these settlements tend to link themselves in a serial fashion. Norberg-Schulz has correctly called this “topological proximity.”3 In general, in Arab-Islamic cities the building mass of the special types such as the mosque, the madrasa, and commercial buildings, is not markedly different from the residential type, as the special buildings are based on the same principle of enclosure and are aggregated with modules no different from basic ones.4 They may be larger, but, with the exception of mosque and citadel, they are never out of proportion, especially in the Maghreb. This is not to say that the notion of a building type - the courtyard, for example - has fossilized over time to become the symbol of Islam;5 there are differences, but within a limited range. Because of the slowness of change in leading types in Arab cities since the fourteenth century and because the range of differences in types was not very great, the settlement appears immutable.6 The other condition, as long as we can speak of urban fabric, is that constitutive typological clarity be accompanied also by the repeatability of aggregated systems or at least that the disposition of the buildings to insert themselves volumetrically in a homogeneous conformation at an urban scale occurs. This phenomenon takes place, even if at times with difficulty, in continuous fabrics, but not in lots of one-family houses. The isolated cottage or house intentionally avoids aggregation because it aspires to the setting of the Palladian villa. In the event that aggregation of such cottages does occur, they are semi-rural in nature and multiply serially, and therefore are neither homogeneous
the urban organism
183 Fig. 124 Synoptic chart of Sicilian settlements with Arab implant: (A) Different textures of tissue shows the phases of the growth of Palermo; the southern tip of the ridge of the Greek city, showing a monaxial scheme (Vicus Marmoreus) from which minor lanes are cut (called the “Phoenician foot”). Note the curv linear strips of the walls adapting to the dictates of topography involving enveloping implant; the regular tissue of the Fatimid settlement called Kalisah; and lastly the rectangular perimeter of the tenth-century walls. Cul-de-sac and courtyard tissues are still legible in the Palazzo Reale and Monte di Pietà quarters; (B) In Mazara the irregular quad corresponds to the Arab town; the neighborhoods north of San Francesco and Giudecca were less affected by seventeenth-century changes and maintained the features of the Arab medieval fabric. (C) The western sector of the old town of Sciacca, including the Rabato, part of the Terra Vecchia, and the ancient ghetto of the Cadda quarter, presents a diffused cul-de-sac fabric. (D) Agrigento, a small Byzantine town on the Girgenti hill, conquered by the Arabs in 828.
formations nor at the level of an urban scale. When the fabric is not made up of a single type, urban nuclei with different leading types can still combine to form an urban fabric under certain conditions. It can happen when nuclei have different orientations but similar functions. A typical case is a rotation caused by topography as in the upper areas of the Algerian Casbah7 (fig. 151), or along the irregular coastline at Cherchell (Algeria) or Dubrovnik (Croatia).8 Another case is produced by the conscious choice to visually contrast a new nucleus with an older one. The (presumed) Roman camp at Aleppo identified by Jean Claude David is an example of this.9 A third instance occurs when urban fabrics are generated by two different superimposed or accompanying territorial grids, as in Sfax, Tunisia, and Florence, Italy.10 Like a sweater knitted in different stitch patterns, urban nuclei can be joined together if they have a similar function but a different weave; the weaves are the result of different socio-political roots. The most obvious example of this is Damascus, where a Hellenistic strigae grid inside the town walls and a Roman centuriae grid outside the walls intersect at a nodal point that becomes the hinge of the two orthogonal tissues. Instead, in the Graziella quarter of Syracuse, due to the complementary land measurements and the homogeneous morphology of the land that did not induce rotations, the Romans fused two geometrical webs together without damaging the older Greek fabric. In Antakya the settlement lies on an axis three kilometers long that constitutes the spine of the ancient city, running at an intermediate height along the slopes of Mount Silpius.11 In the northern part of the present city are traces of a Hellenistic grid 58 × 116 meters that has been noted by Weulersee.12 In the southern
184
after amnesia
part, corresponding to the medieval Muslim city, on the other hand, no attempt to extend the same Hellenistic grid along the monumental street (the present Kurtulus Caddesi) was successful, since in reality the traces correspond to a Roman centuriation. The module found is equivalent to five actus, that is, 600 feet equal to 174 meters, a supermodule however of 58 and 116.13 It is a case of replanning that perhaps can be dated to the Diocletian era, which did not obliterate the preceding Hellenistic fabric. Urban nuclei can even join together when they have different weaves and functions if the typological characteristics are compatible. For example, because of the typological compatibility between special building and residential dwelling, the cohabitation of different tissues with different functions can easily be found in the center of the small and medium sized Arab city. Some planned-building traces are evident in Sousse in Tunisia, and Fez and Meknes in Morocco, where urban fabrics of religious and representative buildings are mixed with clusters of houses belonging to nobles, craftsmen, and the middle class in expanses stretching from the center to the outskirts. Similarly at Tunis, houses and businesses coexist nicely in the mixed-use area around the Zeitouna Mosque, since the special buildings are multiples of the houses. Given these conditions for the formation of urban fabric, there are several modes that allow two urban nuclei to actually join. The modes can be divided into two basic patterns. In the first, the tissues are oriented in the same direction because one territorial direction prevails over the other, as in Antakya. If one direction is blocked for a topological reason - two drastically different ground levels, say, or a coastline - this is usually sufficient to allow a particular axis to prevail, and it is generally one pointing toward a key node in the territory outside the city, such as a landing stage or a train station. This is the case in Monastir, Tunisia. Monastir grew exclusively through the cloning of regular tissues along a single matrix route.14 In the second case, there is conflict among several opposing directions. Typical examples of this are Viana15 and Abaran in Spain, but the case extends also to the two superimposed grids in Sousse and Sfax (Tunisia) (fig. 146.1). The crossed hippodamic grid can be considered a 90-degree variant of this second case. Tissues are generally linked in both ways at different developmental stages of the urban fabric.16 In the structure of the blocks with dimensions of 160 x 420 feet along the east-west route of the central suq of Aleppo, to the west and south of the Great Mosque, it is possible to detect traces of an early plan dating back to the Hellenistic foundation. By interpreting maps at both an urban and territorial scale, it becomes clear how, in Aleppo, there are signs of three different plans in the Roman era corresponding to three different orientations of the building tissue.17 The first plan corresponds to an increase in the urban extension of Aleppo in the same direction as the Hellenistic city; the second plan derives from the necessity to build routes on a regional scale; the third is connected to the actual subdivision of the cultivatable territory and the need to adapt the orientation of the urban and agricultural fabric to the direction of the position of the Qoueiq River, for reasons relating to the drainage of the waters. The first Roman plan is oriented in a north-south direction,
Fig. 125 Plan of Monastir, Tunisia. The parts of the city can be easily distinguished. At the top, the original settlement with its tree structure beginning with the ribat of 796 CE. Following that, two successive modular extensions that keep the matrix route. To the northwest, the Jewish mellah planned separately according to a comb plan.
Fig. 126 Valencia: the morphologically autonomous parts of the city. 1. The Iberian-Roman city 2. Suburbs of the Roman city. 3. The Arab quarters. 4. Rabads. 5. The Christian city with the typical Gothic plots.
the urban organism
corresponding to a grid whose dimensions coincide with those of the centuria or its submultiples, and it can be seen in the urban and agrarian structuring to the south and east of the city: in the width of two blocks east of the Great Mosque, 71 meters wide, a measurement corresponding to two acta, and in the rhythm of a quarter centuria of the routes that articulate the agricultural landscape to the south of the city. The decumanus of this first Roman centuriation coincides with the Hellenistic via recta, transformed in this phase into a colonnaded street, and the cardo coincides with the route to the east of the 71-meter block next to the Great Mosque that reconnects to the polarity of Bab al-Nasr. To the south of the Great Mosque we find traces of a second replanning from Roman times. The road axis which, starting from one of the foundation routes of the Hellenistic city, leads to Bab Qinnasrin, is rotated with respect to the north-south direction at an angle of 18 degrees (fig. 110). This same rotation of the urban fabric is quite diffused to the south and east of the stronghold, but it is legible as well along the road axis that leads to Bab al-Nasr. Beginning with the grid of the north-south plan, it is possible to identify the main axis of a new isotropic system of the structuring of the territory, with a module of 710 x 710 meters, with the hypotenuse of a triangle in a 1:3 ratio, built starting at the intersection between the cardo and the decumanus of the preceding plan and rotated 18 degrees west. Toward the north, this axis coincides with the route that leads to
185
Fig. 128 Synoptic chart of some Mediterranean Arab settlements: (A) Tunis: the drawing shows only the medina, without the extended peripheral expansions. The medina is semi regular implant that is influenced by the geometry of a Roman centuriatio and maybe a small castrum. Note the similarity of this central area with the medina of Sfax. (B) Mahdia: linear implant made out of two fabrics; a grid of three parallel strips at the foot of the hill where the Fatimid castle stands and a successive duplication in the skeleton form up to Place Etjania. The Hajj Mustafa Hamza mosque is on the polar point between the old tissue and its mirroring in the new tissue. (C) Sfax: regular implant influenced by a classic substratum. (D) Monastir: serial implant, easily legible; from an early northern medina the town grew by successive duplications with planned fabrics. (E) Kairouan: Arab formation whose growth is similar to a sui generis geological story; it is possible to distinguish two urban fabrics separated by the Avenue Bourguiba, which in all probability marked the limit of the first city and formed the point of duplication, which marks the start of the second. The first fabric between this line and the mosque has several forks. The second is a more regular fabric. The eccentric position of the Aghlabid mosque is attributed by some scholars to the geographical erosion which destroyed part of the city. (F) Mosul; (G) Aleppo, the Ottoman city. The tissue per strigas of the Greek city is still legible; the subsequent extensions with two rotations are probably Roman. The new radial quarters were just outside the Byzantine gates to the north and west and densified during the Ottoman epoch. (H) Latakia, the Hellenistic implant studied by J. Sauvaget. The grid has deformations particularly in the northwestern areas, perhaps because even during the Byzantine period this area retained a rural character. (I) Acre; (L) Damascus.
186
after amnesia
Fig. 128 Synoptic chart of some Mediterranean Arab settlements: (A) Tunis: the drawing shows only the medina, without the extended peripheral expansions. The medina is semi regular implant that is influenced by the geometry of a Roman centuriatio and maybe a small castrum. Note the similarity of this central area with the medina of Sfax. (B) Mahdia: linear implant made out of two fabrics; a grid of three parallel strips at the foot of the hill where the Fatimid castle stands and a successive duplication in the skeleton form up to Place Etjania. The Hajj Mustafa Hamza mosque is on the polar point between the old tissue and its mirroring in the new tissue. (C) Sfax: regular implant influenced by a classic substratum. (D) Monastir: serial implant, easily legible; from an early northern medina the town grew by successive duplications with planned fabrics. (E) Kairouan: Arab formation whose growth is similar to a sui generis geological story; it is possible to distinguish two urban fabrics separated by the Avenue
Bourguiba, which in all probability marked the limit of the first city and formed the point of duplication, which marks the start of the second. The first fabric between this line and the mosque has several forks. The second is a more regular fabric. The eccentric position of the Aghlabid mosque is attributed by some scholars to the geographical erosion which destroyed part of the city. (F) Mosul; (G) Aleppo, the Ottoman city. The tissue per strigas of the Greek city is still legible; the subsequent extensions with two rotations are probably Roman. The new radial quarters were just outside the Byzantine gates to the north and west and densified during the Ottoman epoch. (H) Latakia, the Hellenistic implant studied by J. Sauvaget. The grid has deformations particularly in the northwestern areas, perhaps because even during the Byzantine period this area retained a rural character. (I) Acre; (L) Damascus.
the urban organism
Bab al-Nasr, and with the matrix route of the Jdeide quarter, and toward the south it is parallel to the axis of Bab Qiinnasrin, the gate that led to Chalcis. The transverse axis of this system, the decumanus, went from Bab Antakia, the gate that led to Antioch. In the urban tissue of Aleppo a third isotropic structuring of the territory and the city is legible, with a module of 710 x 710 meters, whose cardo coincides with the hypotenuse of a triangle rotated about 10 degrees west, and therefore of a 1:6 ratio with respect to the north-south
187
Fig. 129 The walled city of Jerusalem. The plan is based on three successive rotations of the grid.
188
after amnesia
structuring and identifiable with an axis passing through Bab Antakia, tangent on the exterior to the Seleucid walls. The decumanus seems to have been an axis passing through Bab Antakia, near the Roman triumphal arch. This structuring is very evident on both an urban and territorial scale: it traces the rhythms of the agrarian partitions and gives the dimension and orientation of what will be the Mameluk walls of the city, built over Byzantine traces, which are inscribable within a square of four centuria. The technique of successive re-planning of the urban surroundings just seen in the urban tissue of Aleppo seems to have been quite diffused in Roman Syria. In this regard, the urban structure of Jerusalem is exemplary for describing the phenomenon. The planning of Jerusalem in Roman times occurred, on the one hand, by restructuring the existing routes and fabrics, and on the other hand through three partial replannings of the urban fabric that involved different rotations of the building tissue. An early orientation is legible in a north-south direction and coincides with the wall of Haram, the first cardo, the Via Dolorosa and in general the tissue of the present Arab quarter intra-muros, with a decline toward the west of 10 degrees. A second orientation coincides with the direction of the second Roman cardo, laid out beginning with the Damascus Gate (subdivided and tripled in the present suq) and parallel with the Roman Forum (today Muristan) with a decline to the west of 3 degrees. Lastly, a third orientation
Fig. 130 Sfax, Tunisia. The linear and the web-like bazaars, mainly concentrated in the city centre, are pochéd.
the urban organism coincides with a rotation of 3 degrees to the east, corresponding to the orientation of the Roman castrum, whose extension is equal to a quarter of a centuria, whose position coincides with the present Armenian quadrant. This rotation corresponds to the orientation of the colonnaded street in correspondence with the Byzantine extension to the south of the urban fabric18 (fig. 129). The urban nucleus is a module within a larger unit, a superior module of several blocks made up of several building tissues. In a mature urban organism, the fabric is made up not of structures that are relatively independent and exist in typological proximity, but ones that collaborate and become more and more mutually dependent.19 A scalar leap leads to exponential hierarchization, which leads in turn to a maze of hierarchized interrelations. This process of aggregating urban nuclei to form urban fabric happens independently of processes bound to the site and the historical moment. For that reason it is always possible to separate parts of the city that are morphologically homogeneous. That the relation between building tissues and urban fabric is dialectical is confirmed by the fact that the urban fabric is qualified by the attributes of the building tissues it contains. In cities with a layered past, the coexistence of all the conditions is readable in the juxtaposition of different tissues that form morphologically autonomous parts. In Naples, the transition from one
189
Fig. 131 Sfax, Tunisia. The dotted areas mark the mosques, madrasas, and the large funduq. A clear example of the polarity of a religious building and the antipolarity of defensive works. This Tunisian city is an example of the application of the rule of the Jami Masjid centrality according to Marçais. Note that the duplication of the mosque bays has broken the continuity of the cardo.
190
after amnesia
type of fabric to the next is very clear due to the varied topography that conditioned these different tissues. From the Greco-Roman city one moves west to the Byzantine expansion in the Santa Chiara area, to Angevin quarters in the Piazza Mercato area, to the Spanish quarter and finally to the site of the Bourbon royal palace.20 The same is true in Valencia, where in the Roman-Iberian city, the fishbone strip of Arab medina and rabat lined up along the territorial routes and the additions built after the Reconquista, characterized by extremely narrow and elongated lots that are typically Gothic, are all distinct and can be easily identified. According to Paolo Maretto, in Bologna, too, “we witness the Fig. 132 Meknès. The urban fabric and the principal nodes.
191
the urban organism double phenomenon of historical variation and urbanistic structuring of the fabrics: typological variation, from the fabrics crossed at quadrangular blocks, within the ancient Roman perimeter, to the linear ribbon fabrics along the street of medieval expansions; a variation with a fan rotation of the axes of medieval expansion (and hence a varied rotation of the post-medieval fabrics) with respect to the largest two Roman orthogonal coordinates of the cardo maximus and decumanus maximus beginning with the ancient city gates.”21 In the preceding chapters the building tissue was assessed in particular in a serial manner, in its uniqueness, while the urban fabric can not leave out of consideration its greater organicity, in which a centre and a border are always distinguished. To simplify, there is no city without a hierarchy of parts, a notion that implicitly presupposes that of the specialization of single elements.
Fig. 134 Meknès. An urban node. Plan.
Fig. 133 Meknès. Axonometric view of the Jami Masjid.
Fig. 133b Meknès. The central polarity includes (1) Jama el-Kebir (1220). (2) Madrasa Bou Inania (fourteenth century). (3) Madrasa Filalia (fourteenth century). (4) Kissaria. (5) Fondouk el-Sepate. (6) Souq Sebat (thirteenth century).
192
after amnesia
the labyrinth of the hierarchies On the chessboard two grids engage in a contest. In the space and time of the game each changes, both of their own accord and as a result of the changes in the other.The outcome is at the mercy of a very complex mobility, and is so fluid that it is impossible to predict what will happen beyond the third play… We may say at least that it makes little difference which piece is moved first. As time passes the space of interpenetration between the two games grows stronger, and everything happens as if there were a progressive filling out of the concept of determination. Michel Serres Scholarship on the city has only partially represented the complexity of the process with which the city evolves and grows. Michel Serres aptly suggests such a process in his theory of knowledge.22 While his description should by no means be applied literally, just as the fantasy of Calvino’s Invisible Cities is not meant as an urbanistic manual, it does suggest the idea that cities are shaped according to the same web of forces that shapes societies, and that the city is a blend of time and space in a state of progressive coagulation. The main obstacle to analyzing an integrated system like a city is the lack of an obvious point of departure. In fact, the more integrated the system, the more difficult it is to break into simpler pieces. The point of departure for an integrated or organic system can be the tissue, the movement, or the polarities. When discussing the formative effect of movement on building tissues, we observed that the direction of movement was often conditioned by strong or exceptional points that are either hubs of activity or symbolic areas, called here nodes and poles. Conversely, movement can often determine the value of nodes, either absolutely with respect to the whole settlement, or relatively with respect to the dialectic between the nodes. In addition to this formative binary relationship between movement and building tissue are the equally important relationships between movement and polarity, and tissue and polarity. For the sake of simplicity, each component - tissue, movement, and polarity - has been dealt with separately. Tissue and movement have already been covered; we now come to the questions that relate to nodality and polarity. A node can be defined as a moment of concentration in a continuity or a moment of separation between two continua. The relationship between constructed spaces and the general shape of a city is not based on the ordering of simple figures, like mechanically juxtaposed items in a catalog. At the inevitable point where two urban nuclei collide, something much more determined and complex - that is, the node - is constructed. A node is not simply an addendum, but an organism capable of accomplishing a task that is greater in scope than just demarcating space; it is therefore implicit in the concept of building or urban organism.
Fig. 135 The minaret of the mosque of Charabliyin in Fez.
the urban organism
With Serres’s imaginary chessboard in mind we can describe the nodes as the accumulation points of greatest power: “We have a grid of multiple points connected to one another by a multiplicity of branches [the routes]. By definition, no one point is more important than another; the same is true for the routes. However, it is always possible to recut locally strong subsets on the grid as a whole so that their determining strength is greater than the sum of the strength of the individual elements. They are distinct from the grid and have the capacity for more relationships than the primary elements do.”23 In architecture, this increased determination, which we define as an effect of nodality, is not necessarily expressed by a real point, but can also be a virtual encounter between two axes: for example, a Roman atrium domus, or, in a church, the intersection of the nave with the transept, or in a T-shaped Arab hypostyle mosque the axis of central passage and the nave near the qibla, which is marked by a small dome like that of Sousse or Kairawan, and in the Friday Mosque at Damascus, the intersection between the central “triumphal” axis and the longitudinal nave covered by the great dome.24 As a result we expect more legible nodes in the connection of tectonically clear serial structures, that is, systems that are juxtaposed (two joists nailed together or two urban grids) or superimposed (a trilith) in contrast to continuous (arch and vault) or stratified (masonry constructions or a medieval fabric) systems. In a serial urban fabric, as in a building, the nodes are salient exceptions, whereas in a more organic fabric nodes are concealed in a web of hierarchized relationships. A node in an urban fabric can be either a discontinuity in a route - a ford or opening, the intersection of two routes - or a focal point of intersection where two geometric and compositional axes meet. Such is the case for the tetrapilon of Bosra and Palmyra, the four-sided arch in Tripoli (Libya) and the obelisks of a Baroque city. Nodes can also be special building complexes with defining power: think, for example, of the role played by the railway station in a nineteenth-century city. On a territorial scale, nodes are salient topographical exceptions: they can be the settlements themselves, a special locale, a detail of the infrastructure, or points of concentration of resources.
193
Fig. 136 Rue Ibn Khaldun, a matrix route intersected by nodes.. From left to right: Zawiya Macirya and Jama ahl Agadir, Zawiya Jazouliya. Jama Bouakhir. Zawiya Hamdouchiya. Jama Abdellah ou Omar (Thesis project of A. Cutrone, P. Genchi, G.M. Lozupone, L. Massarelli, A. Moccia, D. Panaro, adviser Attilio Petruccioli, School of Architecture, Polytechnic of Bari).
194
after amnesia
A pole is the progressive increment of the concept of node. It is a point at which there is a greater than average convergence of events, the influence of which reaches beyond the immediate surroundings. In a Saharan oasis all paths radiate in all directions from the oasis itself; the Arc de Triomphe in Paris has its convergent radial streets, and analogously polar is the position of the great square Jama al Fna in Marrakesh that determines the radial plan of the city; and the Friday Mosque in Tunis, while not as geometrically obvious, is a pole of many doors all leading to the bazaar (fig. 159.3). Nodality and polarity are concepts that define the quality and quantity of the determinative power of the nodes and poles.25 However, the parallel concepts of anti-nodality and anti-polarity - the counterweights to nodality and polarity in any built object - are not found in an absence of determination but are comparable to the positive and negative poles of a magnet. If nodality includes the idea of centrality, its opposite, anti-nodality, suggests boundaries and limits. At first sight an analysis of cities within Arabic walls shows how urban units of small surface area and the tissues of dense grain are located in the central areas. Nodality, however, does not always coincide with the geometric center: for example, the logical center of a mosque is the intersection in the mihrab of the qibla wall with the central axis. The center of gravity is at the intersection of the central axis and the axis that marks a more internal nave, a point with no sense. In a simple building tissue on a matrix route, the center is represented by the road axis (centralizing axis) and the boundaries of the two edges of the pertinent strip (dividing lines). The center of a Roman castrum, according to Polybius’s description, coincides with the intersection of the via principalis and the via praetoria - the centralizing axes - and the boundary with the pomoerium, delineated by four lines.26 By simply observing urban phenomena we can confirm that the placement of central nodes not only involves the physical shapes of the volumes and the nature of the building types, but also their functions. Nodal functions and services automatically tend to occupy nodal areas, while essential services, many of which tend to be irksome and unhygienic, are placed at the margin.27 In general, shops and businesses prefer the center of a city, and in some tissues they strategically aim for the corner site of a block. Parking lots or the slaughterhouse remain on the fringes. In a Maghrebian city, the Friday Mosque occupies the real geometric center when possible, or at least the center of the bazaar. It is surrounded by functions that do not clash with its sacred nature; there exist successive nuances laid out in a concentric hierarchy for which the closeness of activities is directly proportional to the sanctity of the place, and first there will be vendors of sacred writings, then perfumers and sellers of incense and candles, and, not far away, shops with spices and dried fruit and then the products for which the city is famous and from which it derives its livelihood, such as linen, cotton, silk or jewels, etc.-disposed in progressively secularized circles that reinforce the sanctity of the center. Tanneries and dyeworks are pushed out of the inhabited center to a specialized area.
Fig. 137 Marrakesh. The zawiya of Sidi ben Slimane (fourteenth century). Crosses: oratory at 45 degrees, house of the moqaddem; parallel lines: house for pilgrims; tight parallel lines: school; dotted lines: cemetery. (From G. Deverdun, Marrakesh [Rabat, 1966]).
the urban organism
The determinations are not absolute; the concept of centrality has a subjective component. Kevin Lynch observes that those who inhabit the center of a city perceive it differently from those who live in its suburbs.28 Claude Levi-Strauss demonstrates how the principle of centrality in primitive cultures is a function of an individual’s position in the social structure.29 The dyeworkers in Constantine, Algeria, whose building is on the edge of the ravine, or the leather factory of Fez, whose wells for tanning are along the Fez River that constitutes the caesura between the two main quarters, regard as nodal that activity and that site that are regarded as marginal by the inhabitants. The concept of nodality/centrality is also objectively linked to the scale according to which an object or area is read. For example, a mosque in a neighbourhood can be nodal/central with respect to the neighbourhood and at the same time be anti-nodal/ peripheral with respect to the center of the city. An urban center located in a valley may be a pole in the context of the immediate area, while it is modestly reduced to a node in the greater context of the region. Within the Ayyubid walls of Aleppo of the 12th century, Shmuel Tamari
195
Fig. 138 From antipolarity to polarity: Plan of the axis built on top of the Turkish moat of the Casbah.The Place de la Republique in Algiers, the terminating point of the axis of the French colonial city that includes Boulevard Gambetta.
196
after amnesia Fig. 139 Boulevard Gambetta in Algiers, the line of duplication of the French colonial city.
sees eight urban units with irregular forms, but distinct from the point of view of surface and placement, of which the most important occupies the most central place, Shari al-Kebir. We would tend to assume the absolute dependence of the seven peripheral units from the central one, were it not for the fact that each of them has at least four mosques and a bath, while four of the seven do not have a market. This is a relative dependence: an inhabitant of the al-Akaba quarter can use the services of the nearby alSidla quarter as well as the very central Shari al-Kebir, and consequently mobility is not mono-directional but multi-directional.30 These definitions are historically relative. In the dialectical process by which a city grows there is a continuous shifting of roles between node and antinode, pole and antipole. In the pre-modern city, two building tissues are generally juxtaposed by means of two borders of pertinent strips or two antinodal routes, the fusion which transforms them into a central axis, the nodality, of a larger unit. The extreme ends maintain their peripheral character, but are poised to change this status in future aggregations. Because of the shifting back and forth between node and antinode, pole and antipole, a developed fabric ultimately reveals an absolute center that, with topographical exceptions, confirms the original center, while the outskirts join to form a single outer limit, usually enclosed by a wall. In the ancient city this outer margin marked the beginning of the territory, considered the antipole. On this boundary of the Arab city are placed marginal activities like cemeteries, the villages of those newly arrived and marginalized, the orchards that supply the markets, as well as points of exchange like open spaces for caravans. Also located here is the qasbah, the fortified residence of the governor that does not protect the population but is protected from it, and the idgah, the great space for prayer for the whole community during the festival of Id. The history of the qasbah of Tunis is emblematic: sources tell us that after the invasion of the Hilali the new dynasty of the Khorasanids established itself in a qasr or fortress within the walls of 944, which could
the urban organism be found perhaps to the south of the modern suq el Leffa. The Almohad governors returned to the original model, building the qasbah at the western border of the city, and to underline its complete administrative autonomy, they gave it a hotba or jama’a mosque in 1232. The citadel continued to expand in its present position connecting to the Almohad gardens of Ras Tabia, with a passage protected by walls. The construction of the Bardo Palace further west during Ottoman times signaled the definitive separation of the administrative power from the medina. In Spain, the dichotomy medina/qasbah already prevailed in the 9th century. In Seville during the first caliphate Jami Masjid and qasbah are still together in a central position, and in the Alcazar we still read the quadrangular form of the old Umayyad qasr, of about 180 x 100 meters. Then, in 913, Abd al-Rahman III moved the governor’s residence to the south. This decision is a break with the organic unity of the center of the original medina. After this date the mosque of Ibn Abaddas, the quarter of the silk weavers and the market squares accommodate the religious, educational and commercial dimension, while the political seat is transferred to the future Alcazar.31 In the nineteenth century the railway station was almost always located outside the walls of the medieval city. This pole represented a new dimension of industrial progress, and often a city relocated its center to this area at the expense of the historic center. Naturally, with every shift from pole to antipole, the width of road sections, buildings, and open public spaces varied to adapt to the expanding size of the urban community, and the buildings were specialized to meet the changed needs of the larger system.32 According to this scheme, the unpaved area outside an Arab city, like Tripoli (Libya), for example, may remain the point of exchange between the city and the territory for centuries. It serves as a rest stop for caravans, as the site of the weekly marketplace, and as a place where extraordinary and unpleasant events (such as executions) take place. Initially antipolar, as the urban fabric expands, the site functions as a hinge for the shifting back and forth between two states, by which time the area takes on a new and strongly polar role. A strong specialization of the surrounding buildings usually follows. In Tripoli, the castle, a former Turkish bastion, was restored and became the seat of colonial power.33 Subsequently under colonial rule the Souk el-Kobra (today, Green Square) replaced the medina as the city center only to be replaced in turn by the city hall square during the expansion of the 1930s. In Algiers the phenomenon took place twice for successive doubling of the fabric and is quite visible on a map. In 1830, the city south of Bab Azzoun along the Rue d’Isly grew and thus determined the polarization of the ancient Turkish drainage ditch. Once filled in, it became the monumental Boulevard Gambetta, and the irregular clearing outside the gate became the focal Place de la Republique (fig. 138), where hotels, a national theater, and police headquarters were soon located. The same residential construction was specialized in multifunctional blocks that followed a precise pattern: commercial on the ground floor, workshops and offices on the second floor, and residences on the upper floors, with
197
198
after amnesia
complex distributive solutions such as galleries and the dividing of vertical distributive systems.34 At the end of the century, the walls that had been built by the French were taken down to allow the urban neighbourhoods to be rejoined with the suburbs of Mustafa. This act created the polarization of the long strip of land formerly belonging to the army that ended in a monumental staircase and gardens facing the port. Before long, a host of special buildings such as the Grande Poste, the government palace of Jacques Guiauchain, and the National Library moved to this area. Boulevard Laferriere (today called Le Forum) is the real center of the city confirmed by the central station of the modern underground. Exceptional cases such as Kairawan in Morocco and Assisi in central Italy also confirm this process. In the case of Kairawan, the unexplainable antipolar position of the Aghlabid mosque is justified by the geological forces of erosion that progressively moved the center of the city.35 In Assisi, the basilica of St. Francis - a site of great determinative power reaching out to the entire Catholic world - sits atop the ancient dumping grounds of the city. In St. Francis’s time, it was antipolar and located at a distance from the city center.36 When the center of a city remains in its original location, the exchanges on the periphery that occur with every subsequent expansion are reflected in the original centralizing axis and incrementally increase its polarity. Often this urban axis or supermatrix ends up acquiring a much greater determining power than the surrounding tissue insofar as it reflects the hierarchization of the entire fabric behind it. In Aleppo the Shari alKebir that divides the city width-wise from Bab Antakia to the Citadel is a religious, commercial and government center; with exponentially increasing value little by little as the city extended beyond the limits of the Byzantine city, it has concentrated along its edges not only the markets and the Jama Masjid, but also the secondary civic institutions like the bimaristan. To every increase in the surface of the city there corresponded an increase in the volume of the institutions and the quality of the architectural structures, up until the monumental redesign of the covered bazaar in Ottoman times.37 Massachusetts Avenue between Boston and Cambridge, the Qasaba of Cairo and the Divanholu, the great artery in Istanbul which, beginning at Hagia Sophia, connects the Great Bazar to the mosque of Shehzade and Fathy38 excuse us from a more in-depth description. The opposite can also be true when the mechanism for shifting from pole to antipole and back again is not activated. In this case the city grows in a straightforward arithmetical way and remains a large collection of urban nuclei strung together without hierarchy, as in the above-mentioned case of Monastir. In a sense, special buildings follow the same dialectic. In Arab cities some buildings - the mosque and the suq, for example - belong in the center because of their nodal nature, while others, such as industrial buildings, are gradually pushed to the outskirts as the city continues to grow. The former tend to highlight the center with a concentration of mass whose impact is multiplied by the rotation of the tissue in the
the urban organism
199
Fig. 140 Kairawan. Measured architectural survey of the ground floor of the city. The mosques are marked in dark. (redrawn on the basis of the survey of P.Jervis and P. Donati)
200 after amnesia direction of Mecca. The anti-poles instead have a tendency to distinguish themselves from their surroundings by an enclosure, and generally occupy a very large area, the location of which is influenced by both the low cost of land and the types of activities that it attracts. One thinks of the areas of the pottery sellers at Bab Fetouh in the Andalusian city in Fez. Once the roles of node and anti-node are exchanged, this area becomes the object of division and transformation. In the West and in Italy in particular, this phenomenon is typical of medieval convents. Initially peripheral, they provided the stimulus for the formation of autonomous villages. As the city encroached on this area over a period of many centuries, the modification of the residential tissue of the village reached its limit and often all that remains of the convent is the church. It is relegated to the status of neighborhood church and all the convent’s production and storage annexes are given over to the construction of housing. Even the zawiyas of Maghrebian cities are found at the edges of the urban fabric or in the strip immediately inside the walls. However, once they collide with urban expansion, rather than become hierarchized they tend to be camouflaged in the new residential tissue as in
Fig. 141 Special buildings in the fabric of Damascus, south of the Jami Masjid.
the urban organism Kairawan (fig.100.4), Essaouira or Gadhames in Libya. The greater determining power of central poles that goes hand in hand with the formation of the urban organism is manifested differently in serial and organic areas. The intrinsically serial special buildings of Maghrebian cities usually grow through successive duplication based on population growth. The case of the mosque/cathedral in Cordoba is typical. It grew at the expense of the adjacent residential tissue through two successive duplications between 785 and 987, and, with its 130 x 180 meters at the end of the western caliphate, it was one of the largest prayer rooms in the Islamic world,40 and albeit at a different scale the hotba mosques of Sfax and Hammamet in Tunisia.41 Because of its pronounced organic quality the Ottoman mosque, as opposed to the Arab and Persian one, can not grow by doubling and therefore must be demolished and rebuilt.42 In the West the opposite occurred; the central pole tended to limit its own powers of encroachment and even complement the acquired organicity through the complexity of its spaces and growth in height. In the Arab city, between the high levels of nodality, services with a high degree of local influence join smaller residential modules. The marketplace and the city share this dialectic: the bazaar is central, and while it offers all types of goods, including precious gems, each quarter has its own commercial node that supplies basic foodstuffs, and at the very
201
Fig. 142 Damascus, Syria: the grouping of alignments identifies two regular patterns with minimal rotation. The first is aligned with the great decumanus called the Rue Droite, an ancient ridge path descending from Jebel Kalabat el-Mezze, and establishing the orientation of the citadel and of the temenos in Jupiter’s temple; the other one, oriented exactly according to the cardinal points, determined the pattern of the residential areas. Note some leftover paths and diagonal routes within the fabric, suggesting open spaces in the temenos (today’s Harat an-Naqqasat), or in the Christian quarter (where Sauvaget hypothethically located the agora).
202
after amnesia
least there is always a bakery to be found. In Essaouira the bazaar extends beyond the intersection of the main roads in the street. In addition, little clusters of two or three shops selling necessities are found in the heart of the residential quarter.43 The same holds true for religious services: though theoretically there is only one mosque/cathedral per community of worshipers, each quarter has its own and a network of chapels spreads out to serve even the smallest of neighbourhoods.44 Over time, the tides of history and the growth of a city together bring about a concentration of services, eliminating the smallest, and the specialization of the remaining services which increase in size in order to carry out their new functions. This is statistically proven for Italian parishes but is less well supported for mosques.45 The Arab city accepts the idea that there will be more than one Friday Mosque as it grows, either because the size of certain urban organisms is such that the global organism has already experienced a relative loss of identity with regard to subcenters,
Fig. 143 Miliana in Algeria. The ridge that generated it.
Fig. 144 Constantine, Algeria: the ridge leading to the plateau along which the city developed formerly had a primitive acropolis (today it is a military area).
Fig. 145 (Right page above) Sfax, Tunisia. Grouping the various alignments identifies two main grids which were the genesis of the city. The larger grid corresponds to the territorial division of Roman origin. The second, smaller grid delineates an important rotated structure that could have been a castrum or a second planned extension. Note how the curvilinear routes avoid obstacles (probably special buildings of classical origin). Fig. 146 (Right page below) Sfax, Tunisia.The superimposition in black represents the trace of the contours which allow us to identify where the main ancient river was; it has now completely disappeared.
the urban organism
203
204 after amnesia or because of the immigration of some group, such as the Andalusians to the Maghreb in several waves beginning with the reconquest, who seek to preserve their own cultural identity. So in Essaouira, a medina of 60,000 inhabitants, there are four mosque/cathedrals; in Tunis at least twelve; and in Damascus there are about ten around the Great Mosque only. The case of Tunis is yet again an emblematic example: the Jama’a ez Zeituna was built in 732 or 734 and further enlarged in 864. The Zeituna remained the only congregational mosque until the 13th century, while Sousse, Sfax and Kairawan each had two or more. According to the Malikite rite, this would indicate how access to the mosque by the entire population of the medina was not blocked by walls at the time. The construction of a hotba mosque in the citadel immediately after the consolidation of the Hafsid domination confirms its autonomous role and that of its inhabitants. Toward 1250 the third hotba mosque was consecrated in the suburb Rebatt el Marr and a fourth connected to the population of the Barrani in 1283 outside Bab Behar: the Zeituna al Barrani. At the end of the 13th century with the growth of the two suburbs of Rebat Bab Jezira and Rebat Bab Suwaiqa there were already six hotba mosques. In 1451 the suburb of Bab Sa’adun was incorporated with the creation of the new Jam’a Sidi Ja’afar mosque. This southern suburb grew so rapidly that an eighth was necessary between 1451 and 1535, promoting to the level of hotba the Mosque of the Ring that already existed outside Bab Jedid.46 Even in precolonial times the mechanism for determining nodes and the bonds between parts and nodes intensified to the point that in the city-metropolis the center of the medina could become a predominantly special tissue, with mosques, madrases, caravansaries, urban ribat, monumental suq, as in the case of Tunis (a happy medium) and especially in the case of Cairo,47 and above all Damascus, in the quarter to the south of the Great Mosque (fig. 141), and Jerusalem in the strip, a block deep, that borders the Haram to the north and west.48
the urban organism
205
interpreting the urban fabric Only chaos can produce an illegible structure. When an urban fabric appears to be confusing it is usually because we are either insufficiently prepared for reading complex structures or because we are dealing with characterless, disintegrated forms such as the modern outskirts of a European city or the seemingly unending American suburb.49 This can also be true for Mediterranean cities. Beyond morphologically intact medieval walls or nineteenth-century quarters, there lies a hybrid landscape of architecturally heterogeneous objects, whose functional and formal logic responds only to internal laws.50 The laws that govern the relations between the objects are so reduced that the interstitial spaces, and consequently the context, become a no-man’s land. The opposite is true, however, for the continuous urban system of a traditional city. In spite of its complexity, one is able to both reconstruct a hidden formal structure and the relational laws of the components and untangle a maze of interdependencies. The reconstruction of the structure requires a series of steps, each of which functions as a filter to allow the individual components to be separated and the parts to be divorced from the organism as a whole. The steps of the reconstructive process are, in order:
Fig. 147 Algiers. The ridge routes of access to the Casbah from the hinterland. Note the position of the Turkish fort, later called L’Empereur, on top of the route. Beginning with the citadel, two secondary ridges point toward the route of the bottom of the valley between Bab el-Azoun and Bab el-Oued.
206 after amnesia
1. Distinguishing the territorial routes that connect the city with the major regional poles. On a city map making the typical error of viewing the urban system as a closed body inside a circuit of walls and leaving out the surrounding geography would be like looking at an abstract black-andwhite graph. Instead one should topologically insert the construct into the territory. These routes are easy to make out when they trace original paths, give or take some technical corrections, but are much more difficult to locate in the urban outskirts where they might have been erased by the extensions of the suburban quarters. The routes of a territory are its most permanent marks, since they lead to cities which are not easily moveable. Territorial routes usually meet inside the walls in the structuring routes of the ancient fabric and lead to a major intersection. In spontaneous systems this point coincides with the founding nucleus of the city. The diagonal cut-unusual in a relatively regular grid-that crosses the medina in Tunis from the Jama’a Zeituna up to Bab Carthagena can be understood, for example, through the persistence of the ancient route of Carthage. Not every territorial route is able to meet the demands of modern automobile travel, however, so the contours near the walls of a city instantly reveal the now-abandoned ridge routes that made the founding of the primitive settlement possible. If experience is any indication, every urban system has at least one entry ridge. Miliana (fig. 143), Constantine (fig. 144), and Algiers in Algeria; Latakia and Damascus - where the via recta in the Acts of the Apostles is nothing more than a raised route on the Barada plain - were all generated by ridges. Even Istanbul’s main axis of penetration, then the ancient Egnatia road, is a long ridge that culminates in the Topkapi promontory.
Fig. 148 View of Algiers in the sixteenth century. The Casbah is shown as a continuous fabric without the nodes. Anti-nodalities, on the contrary, are emphasized, such as sea defences and the little island of the Admiral.
Fig. 149 (right page) Algiers, the Plan Pelet of 1834. Three main types of alignments can be seen: (1) aligned to the rue Bab Azzoun(A-A); (2) part of the grid in the flat area de la Marina; (3) generated by rue Mohammed Azzouizi. The chronology of the alignments is uncertain. (D) Present position of the Jami’ Masjid. (E) Location of the ancient Phoenician harbour. (F) Citadel; 1. Bab el-Oued; 2. Bab el-Azoun.
the urban organism
207
208
after amnesia
Fig. 150 Urban fabric of the north part of the casbah in Algiers (area bounded by Boulevard de la Victoire, Boulevard Gambetta (today’s Ourida Meddad), and the Rue de Verdun (today’s Abderahmane Arbadji). Note the alignments perpendicular to the contours defining building plots of approximately one actus. The alignments correspond to the ascending routes, and rotate according to the contours until they become orthogonal to the earlier orientation. The dashed triangles represent the leftover areas absorbing special topographical conditions. The routes along the contours and the stepped paths orthogonal to them are shown in black. The only two diagonal routes converge near the streams in ancient nodes (they were probably fords).
Constantine - one of the most spectacular and overlooked sites in the Mediterranean - is also worth mentioning. Sited to take advantage of a unique natural gift, the city sits directly on a deep, sharp gorge cut by the river Rhumel, whose ravines naturally protect the plateau. Centuries ago, a clearly defined ridge led man to the high part of the plateau to find an acropolis that later became home to both the Numidians and Romans. From the acropolis a regular grid was later traced to cover the whole plateau. Confirming the strong permanence of the place on the acropolis, succeeding one another in time, were the qasbah, the French barracks and the administrative structures of the new independent republic, all directional activities with a strong polar accent. By descending the ridge route that starts at the Sahel and forks at the top of today’s citadel, one can see how the Casbah of Algiers evolved from the top down by a system of routes riding the secondary ridge and meeting the Phoenician city on the line of Piedmont (the present segment Bab Azzoun - Bab el-Oued) that later became the backbone of the classical city (fig.147).
the urban organism
All promontory settlements follow this formula (fig.154). This step allows us to determine the original skeleton of a city and the original iconography imprinted by the territorial morphology. One example can stand for all of them: according to Levi-Provencal, in the territory of the present-day Fez el Bali, Idriss II founded the city of al-Aliya in 808, after having de-Berberized the area.51 The second nucleus, Madinat Fas, was developed beyond the Wadi ‘l-Kebir as an extra-urban suburb, after the massive immigration of five-hundred Andalusian families in 817. A simple study of the positions of the three gates of this second nucleus overturns Levi-Provencal’s thesis and demonstrates how the three large territorial routes converge toward them: respectively from Rif on Bab Abi Sufyan, from Tlemcen on Bab al Kanisa and from Sigilmasa on Bab al Fawwara. Madinat Fas is therefore the original nucleus, while the nucleus of al-Aliya separated from commerce reveals its origin as a court city founded only later. 2. Overall reading of the forma urbis. All formalistic approaches to urban historiography have proved to be of little use. Tricart’s generic division into homogeneous and heterogeneous cities and then heterogeneous planned
209
Fig. 151 The casbah in Algiers: in the quadrant between Boulevard de la Victoire and Boulevard Abderrezak Haddad. The grid starting from Rue Mohammed Azzouizi determines the transversal street pattern. The building tissue in the blocks bounded by Rue Sisi Driss, Rue Smala, Rue Mrouche Mahmoud, and Rue Nfissa mediated between the two grids at 45 degrees to each other.
210
after amnesia
Fig. 152 Algiers: the block bounded by Rue Sidi Driss, Rue Smala, Rue Mrouche Mahmoud, and Rue Nfissa, plan.
Fig. 153 Algiers: the block bounded by Rue Sidi Driss, Rue Smala, Rue Mrouche Mahmoud, and Rue Nfissa. The first is a matrix route corresponding to one of the two ancient ridges cutting the casbah; the second and third are two planned routes intersected by the first; the last is a connecting route. On the matrix route, note the substitution with apartment houses of the colonial period.
and heterogeneous in their topological development is useless.52 More useful are those classifications in which the forma urbis has been connected to the historic components of the plan, as in Lavedan53 or in Piccinato.54 The latter, in a valuable little book on medieval cities, used a formal classification in which some categories like “herringbone” or “chess-board” have an intuitive correspondence with the plan, and others like “fused” or “fan” are purely formalistic, generating a system of non-equivalent elements. More useful are the systems of classification of Conzen55 and Reps56 that work on the identification of phases, in which the periods of expansion furnish the components of the plan that are to be isolated. Here, I propose a more effective method that goes back to Muratori’s
the urban organism theory of the territory,57 evaluating the forms of the sites through crossed parameters referring on the one hand to the morphology of the territory in four synthetic meanings of ridge, cross-ridge, valley bottom and plain, and on the other hand to the three geometric essentials of point, line and grid. As can be seen in fig. 154 twelve categories are obtained that we could define territorial types with a low level of typicality: punctiform on the ridge, punctiform on the cross-ridge, punctiform in valley bottom and punctiform in plain (absolutely theoretical, in fact the urban nucleus coincides with the building unit); linear on the ridge (in which the nucleus of the promontory is an important typological variant), linear on the cross-ridge (a common type inasmuch as it is coherent with both the movement along the contour line and the growth of the aggregate), linear in the valley bottom (in the two forms of site on the piedmont and embankment site) linear in plain; reticular on the ridge, reticular on the cross-ridge, reticular in valley bottom and finally reticular on the plain (the most used whenever this territorial morphology favours isotropic geometry). In the complexity of the processes of growth the twelve types identified are applied only to very serial elementary nuclei, where in every minimally evolved settlement more nuclei are legible that belong to one or the other type. Thus, with the more opaque parts of the plan removed, like the tissues of obstruction, it could happen that a small hill settlement has an original linear nucleus on a ridge, with a successive linear expansion on the cross-ridge and another one lower down, reticular on a false level of a valley bottom. It should be noted that the special buildings and polar activities are always located in the points of tangency between nuclei. If, on the one hand, the classification of a site using this method seems more complicated than a simple formal label, on the other hand it guarantees an evaluation of the forms responding to the structure. It should be added that a trained eye is able to classify a site using this method not only from the design of the plan, but also from a direct look at the site. 3. Distinguishing morphologically homogeneous urban parts. It is always possible, on the surface, to separate areas of homogeneous urban fabric by the geometry of the layout and the grain of the tissue. However, in order to understand the slight variations that exist within a given quarter or the relationships between distant quarters, one must look at a cadastral map or measured architectural survey to identify orthogonal systems and their internal laws58 (fig.142). The aim of this second step is to separate the different warps into groups and establish a first chronological order based on the principle that there is no territorial, urban, or building structure whose planning characteristics do not reflect the technological and social advances of its time in history. To be of use, the interpreting should be done simultaneously on a cadastral map reduced to 1:2000 for the urban scale and on a building survey of 1:500 for the building scale.59 In the orthogonal systems the most promising lines for a reading are long ones, though this almost always means line segments that can be connected if they occupy the same position on a plane. These long lines might be a sign of systems that are orthogonal on a higher level than a quarter and involve the entire plan of a city. The final product is a synthetic
211
212
after amnesia
and immediately legible chromatic table that indicates homogeneous parts. Warps fall into four major groups: the crossed modular fabric; the monodirectional fabric; the fabric that pivots on large founding ridge routes; and the mediating diagonal fabric. The first group includes the Roman foundation fabric derived from a castrum or a municipium, and the second includes the subsequent expansions where the predominant direction is perpendicular to the contours.60 They are usually used in a situation in which land is divided into parcels like terraces for farming, and that later on become building lots. The third group includes warps generated by the steepest and oldest access routes to a city. Branching out from these are system routes in the form of a comb that are parallel to the contours of the land.61 Finally, the fourth group includes fabrics that fill in the bottom of a valley whose diagonal position is determined by the need to connect two fabrics that go in opposite directions. These four types of warps coexist in hill towns with a classical past. The next step in the reconstructive process is to distinguish the internal laws of each urban part and understand the connections between them by using the information from the analysis that relates to modulation, orientation, and dimension. Discord between structures of different scales is possible at this point in the reading, if the laws of the architectural-building scale are different from those of the tissue or urban organism. For example, in a Maghrebian fabric where smaller building units are incorporated into large bourgeois houses without first making considerable changes in the new unit, there are no obvious differences in Fig. 154 The twelve territorial types. The double entry grid integrates the elementary territorial morphologies with the settlements and routes.
the urban organism
213
the design. Only an architectural reading of the doors and openings can separate these two chronologically different phases. 4. Identifying the natural base of the site. This is a crucial step and involves superimposing the chromatic map on the contours. One advantage of this is that it translates a flat geometrical scheme into a threedimensional one. It also helps retrieve natural features, such as ditches, sand banks, diverted waterfronts, landslides, dikes, rock seams and diverted rivers that have been obliterated by human intervention and appear as interruptions in the continuity of a fabric (fig. 146.2). In uneven terrain the grouping of homogeneous marks shows up like the islands of an archipelago. In plain settlements (a perfectly level plain naturally being an abstraction, the fact that the attraction of the territorial nodes is always unequal and varies in weight over time), the warps tend to follow one another theoretically without any rotations. Fig. 155 Diagram of the routes of Damascus: the culs-de-sac are solid dark; the two matrix routes are shown with continuous lines; the main planned routes are shown with dashed and dotted lines; dotted lines identify the secondary paths serving the residential areas.
Fig. 156 Damascus. Neighbourhood units served by culs-de-sac and not accessible to outsiders.
214
after amnesia
5. Identifying polarities, antipolarities, intermediary hierarchies, and rehierarchizations. This step and its theoretical implications have been adequately discussed in the previous paragraph. We can add the usefulness of a representation that makes evident all the special building compared to the entire built environment, as in the renowned plan of Rome by Giovanbattista Nolli (1734), made famous by the school of Colin Rowe at Cornell University.62 With great graphic effect but little usefulness for interpretation are instead the plans of the fabric in which the urban solids or voids are alternatively painted in the background in black. 6. Identifying preexisting layouts through the enduring clues of geometry. While inscriptions can be moved from one place to another and historical documentation falsified, geometries and their measurements paint a clear and unmistakable picture of the actions of a particular society; they are its signature, if you will. From the pattern of parallel lines on the chromatic table one can rather easily deduce a common multiple - the module of an eventual geometric grid, or perhaps one of many - and this helps us to date the urban system. This step is particularly important in the study of Mediterranean fabrics because many of them, often unbeknownst to their inhabitants, contain an extensive and cumbersome number of classical remnants.
Fig. 157 (left) Diagram of the routes of Sousse, Tunisia (same notation system as figure 155). (right) Neighbourhood units served by culs-desac and not accessible to outsiders.
the urban organism
215
one city does not hide another It is a given in urban historiography that periods of growth leave permanent marks in the plan of a city. In his studies on Paris, Marcel Poete first formulated the universally accepted thesis of the permanence of plan.63 Pierre Lavedan later verified the assumptions in a more extensive study.64 This general idea can be extended from the shape of a layout to all building scales. If a city grows in fits and starts, alternating between moments of decision-making and individual initiative, the imprint of both is mixed in the urban fabric, and a philological reconstruction should bring them to light. The object of this type of analysis is usually not the archeological site, even if we refer to it sometimes for the sake of comparison, but the living city, in which civitas and urbs - the social and the physical city - are linked. If we were to focus exclusively on archaeology, and if we were able to uncover the mass of mutations layer by layer, then a good archeological survey would be sufficient and all the geometrical and typological knowledge in the pages of this book would be superfluous. Alas, the living city does not allow the archaeological exploration of its body except in rare cases such as in Murcia, Spain,65 or for example when large-scale public works required extensive demolition, as in the fora in Rome, and within a limited area.66 The archaeologist who wishes to explore buried layers without excavating has no choice but to collaborate with the historian, the interpreter of written documents, or the typologist, who conducts a sort of non-intrusive archaeology by deciphering the marks of the forma urbis. With the help of examples and constant reference to the drawings, some important principles allow us to uncover extant remnants under the thick crust of an urban fabric, but only on condition that the older planned structures have been recovered gradually and spontaneously.67 Where a city has been abandoned and its structures recovered only after a long period of time, the harmony between the traces will be very limited. Today’s Mediterranean city as we know it was essentially formed in the Middle Ages. Subsequently an extraordinary architectural solution was achieved by means of a series of procedures that closed off the network of relationships, filled voids, and increased the specialization and concentration of the nodes. The plan became increasingly more organic but the iconography remained substantially unchanged.68 The iconography we are examining, however irregular in form, contains fragments of regular geometry that upset the homogeneous growth of the spontaneous city. These recurring signs behave in a typical and systematic way, and can therefore be isolated and linked in logical relationships: they can, in other words, be interpreted. In the flow of history they represent the pivotal moment of either individual or collective decision-making that leads to the founding of an urban organism or to its extension on the basis of geometric control, a fundamental aspect of any city plan. Each plan is defined by a desire for geometric rationalization, but it is rarely possible to preserve the crystalline form through time because rule and transgression always coexist in the city.
Fig. 158 Hypothetical reconstruction of the building tissue of the casbah neighborhood in Essaouira. From a primitive settlement of south-oriented courtyard houses with double layout to a ksar with covered passages. 1. Palace of the Sheik, now demolished. 2. Jama Masjid. 3. 4.5.6. Larger areas occupied by the Grandees and further encroached and subdivided. Note how the present place de France is sited on the old route aligned with the mosque.
216
after amnesia
A 1:10,000-scale map shows winding extra-urban routes that connect territorial routes. The walls of a city - either torn down during a growth phase or enclosed within the fabric - are revealed through an intricate tangle of curves. Extremely convex routes that lead away from the city may meet some concave ones at some point; they are deflected as if they had hit something, and pass tangentially. The types of structures that usually close a system are located near the points of deflection; the union of the corners restores the total or partial perimeter of the original urban circle. The Arab village that encroached on the sanctuary of Ba’al in Palmyra demonstrates a series of typological behaviours straight from a manual, from territorial routes, which, rejecting the corners of the enclosure, follow a curvilinear route toward the openings, to the processes of the encroachment of tissues.69 The rounded bulges on the smaller scale of the block, which we call building increments, are also helpful in determining the obliterated ancient perimeters of large urban complexes. In interpreting, it is important not to be distracted by convexity, but to hunt for the original alignment inside the built object on the principle that internal tissues are better preserved, and extremities--or the points where the increment begins and ends--are usually attached to the original alignment. Sauvaget first applied these criteria to the reconstruction of Greek Damascus and found that the broken shape of the northern ring was actually a medieval development. The primitive rectilinear outline can be found, according to Sauvaget, on the segment that joins the extant north gate and the St. Thomas gate.70 We find analogous additions in the other three sides of the city that tend toward zero in correspondence with the western and eastern gates, and set back with regard to the southern corners; the walls of the medieval city are therefore curved and their points of flex pass through the two gates of the decumanus, which are therefore in situ. The present bulwark of the eastern gate develops along a rectilinear line perpendicular to the direction of the decumanus. This extended bulwark ran through the northeast corner tower of the ancient city. This same technique of studying an imprint or impression in order to discover the original shape - for example, in Pompeii archaeologists pour cement as if in a cast into empty spaces and people and animals suspended in time miraculously come to life - can be applied on a smaller scale. In this way military structures emerge from within dense building tissue, detectable by the presence of towers or bastions that have been incorporated into homes. Similarly, large palace enclosures, special buildings and fortress-like compact residential tissues emerge, as in the Tunisian block bounded by the rue Tourbet el-Bey, rue du Persan and rue des Judges. The eastern wall of the temenos of the Roman temple in Damascus is a germane example of archaeological evidence supporting the typological principle of alignments. According to Sauvaget, the direction of the eastern enclosure is revealed by joining two fragments of the door with the massive remnant of the wall. The archaeological discovery of sixteen in situ shafts of columns on the inner street, which determine the width of the portico parallel to the external wall, corroborates the find. The bend in the road that enters Diocletian’s Palace in Split from the
217
the urban organism
1
3
2
4 Fig. 159 The flexes of the curvilinear paths identify the geometric structures according to tangent points. Note the different modes of encroachment onto the interstitial spaces. (1) Karawiyyin mosque in Fez, Morocco: The entire southern triangle is filled up with services for the mosque/madrasa like the library, and to the east the tangent Street has caused a deformation of the bays of the hypostyle hall. (2) Rue de la Kasbah in Tunis:
note the path terminating in the block. (3) Zeitouna mosque in Tunis: note the rotation of the qibla wall mediating between the real direction of the Mecca and the urban grid oriented according to the four cardinal points. (4) Jami’ Masjid in Sfax, Tunisia: note the two tangent streets, the diagonal one that has generated the encroachment tissue behind the qibla and the street tangential to the minaret that has determined the curvilinear deformation of the last nave of the mosque.
218
after amnesia
eastern gate bends in two places, and there are enlargements on either side. The original road section is easily reconstructed at the point where they are tangent. The ancient city gates, even if obliterated in a later doubling of the fabric, are almost always detectable because of the marks left by transverse and radial routes that meet there (see the example of Seville in figure 89). At times these imprints correspond symmetrically to others inside the gate. This suggests that at a certain point a semi-rural open space between the constructed border of the city and the walls was devised to ensure continuity, and the routes were able to cut through it, freely moving to the gates as in the Anatolian city of Urfa.71 Converging roads always indicate a forced point of passage or a constriction, like a bridge. Curves in the urban fabric, besides deviating to avoid obstacles, also form either junctions, which rectify sharp differences in the level of the ground where vehicular traffic is necessary, or else arrange lots in areas on an incline that is parallel to the contours of the land. Istanbul offers an example: on the incline that leads from the Grand Bazaar to Eminonou, the regular tissue of the domus behind the Egyptian Bazaar alternates with diagonal ones that form triangular and pentagonal blocks. The latter are home to the large Ottoman hans which adapt better to irregularities thanks to their extremely serial, and therefore malleable, layouts. An example of the second case is found in the universal method of arranging a fabric on terraces where there is an absence of vehicular traffic and with minimal movement of land. The junctions between one ground level and the next are resolved with ramps perpendicular to the contours of the land, as in the upper part of the Casbah of Algiers between boulevard de la Victoire and rue des Pyramides (fig. 150). Not all curvilinear marks refer to deviations and obstacles. The hemicycles of classical special types such as the circus, theatre, odeum, naumachia, and amphitheatre are obstructions so easy to discern in the fine weave of a medieval tissue that they are almost always easily identified in the reconstructive maps made by urban historians.72 Once large open spaces and courtyards in the interiors of special building no longer serve their original public function they are rapidly metabolized by the inhabitants. A society will defend a public space only if the site continues to be symbolic or retains a special function, though it can be different from the original one. Often a new tissue forms on one or several diagonal routes. By definition, these then act as new matrix routes that in turn generate systemic and connecting routes, and then a spontaneous tissue that fills in all the available open space. As a consequence, in the illustration of warps the presence of special serial buildings like basilicas and fora is signaled by a sudden densification with respect to the surrounding residential tissue. Hence it is not paradoxical for such grand classical spaces as the fora or the vast interiors of the horrea to be today the most densely filled in with a tissue that is rotated with respect to the outside perimeter. Think of the forum in front of Pompey’s Theater in Rome for example, or the Greek agora in Damascus that Sauvaget inserts into the only zone where the classical grid was obliterated by a knot of crooked little streets.73 The shape of Diocletian’s circus in Rome
Fig. 160 Kairawan, Tunisia. The building pattern to the right shows that the street once continued here and has now been encroached upon.
the urban organism
219
Fig. 161 Algiers: the block between Rue Lounas Mustapha and Rue Lalahoum showing the structure of a Roman theatre.
is perfectly preserved because it was first turned into a marketplace for the city and later into the splendid Baroque square known as the Piazza Navona. In the West, where cathedrals often reused classical foundations, several fora were preserved by becoming the churchyards of cathedrals. The agora of Aleppo and Diyarbakir were incorporated into the courtyard of the Jami Masjid and, as a result, their character as a public open space was preserved - though in a wholly different context. The complementary phenomenon to the infill of the open spaces that joins the geometric corruption is the distortion of the Hippodamean grid. At some point a street might be obstructed by a private citizen or only in part occupied by the increment of a row of houses. In the first case, this event interrupts the flow of traffic and involves the deviation that will be diagonal if in an unbuilt area, or stepped if it meets a fabric of regular blocks.74 In the second case we have the formation of those long curvilinear fronts. A second partial form of obstruction is seen in the jutting in front or building extensions in the public street that in time generated the gently winding medieval streets. Generally they have bulging forms since the addition occurs more easily in the center of the block or built row and rarely at the corners, so that the additions are curves that tend toward zero at the corners. Nevertheless the original alignment of the classical street can be retrieved since it is preserved in the inner tissue of the urban block and since the starting points of the curves coincide with the corners of the original classical block. This also holds true for the most intricate and stratified fabric like the “organic” Arab one. The present layout of the Casbah of Algiers is the complementary phenomenon that points to geometric corruption like the distortion of the Hippodamean grid.
220
after amnesia
The present layout of the Casbah in Algiers is an interesting topological structure. An analysis of its warps reveals three main networks, probably not from the same period, separated by three routes (fig. 147) laid out beginning with the three routes that make up the plan of the city: the two main descent ridges and the Bab Azzoun- Bab el Oued piedmont. The medieval fabric covering an ancient classical layout appears deformed because of two complementary actions: infill of the open spaces and deformation of the geometry of the plan. The reconstruction requires a careful separation of the two phenomena in order not to confuse the results of one with those of the other. The opposite can also occur. The classical layout of an area may persist despite the adversities of history and the loss of memory. In Hammamet, Tunisia, a regular layout, presumably medieval - for there are no historical references to it before the thirteenth century - contains the original core of a small settlement based on a 2 x 2 actus module.75 This in effect demonstrates that the system was derived from the ancient Roman practice of dividing land into plots. As the figure demonstrates, the town grew by simple additions that did not deform its morphology. The mosque and the small bazaar were added at the northeastern margin of the existing village, then a castle to the north, and subsequently more tissue, all modular. In the case of Hammamet, with metrological skill it was easy to reconstruct the original land division and isolate the original Roman nucleus because of its serial growth. Metrology is the unjustly neglected supporting tool useful for successfully analyzing the substratum of geometrical structures. Measurements are the calling card of a people. Modules and ratios cannot lie; they are like the signature of an artisan. Few medieval tissues in the western Roman Empire use measures that are the multiples of the actus76 because the measurements themselves were victims of the political decline and fall of the empire beginning in the fourth century. The renewal of building in the Middle Ages not only took place using entirely different types of measurements, but measurements without any standard. In an exaggerated display of provincialism,
A
B Fig. 162 In the curvilinear paths of a medieval town, we can still identify regular structures absorbed by the tissue. Two examples of small settlements of Roman origin: (A) Albiolo, (B) Formia; and of a double Byzantine castrum; (C) Sarzana.
C
the urban organism measurements even differed from one city to the next. In some cases metrology is the only instrument we can refer to in order to date a remnant of history. In Tripoli, Libya, where the boundaries of the imperial Roman cities have never been clearly defined in relation to the subsequent Arab and Ottoman fabrics, the cautious measurement of a block presumed to be on the ancient decumanus (today Sciara Quscet el-Seffar) found it to be a block of two actus with a further 30-foot subdivision, reinforcing the hypothesis that the five equal blocks south of the decumanus are also part of the imperial development in the Roman town (fig. 165). The situation was different in the East. Civic continuity and an economic boom in cities under Byzantine dominion favored uninterrupted continuity at least until the Ummayad period. Because Byzantine metrology adopted values similar to Roman ones the measurements were very accurate.77 This means that, besides archaeological and historical information, a new body of knowledge could be added to the present one simply by paying attention to the measurements of a building and looking at its marks in the plan. New venues for research are opened by changing the terms of a debate. The study of the urban organism using the original Islamic measurements can shed new light on the historical facts and definitively address our many presumptions.78
221 Fig. 163 Palermo. Along the traces of the ancient Sebeto river (today’s Corso Venezia), a block on the matrix route hides a portion of the ancient urban walls in its interior.
Fig. 164 Aerial view of Tangiers, Morocco. A portion of the ancient city walls embedded in the urban fabric of today
the urban organism
223
the city: neither spontaneous nor created E. Pauty’s well-known article, “Villes spontanées et villes crées en Islam,”79 is identified with the debate that is at once ethnological and ideological, and that has from time to time linked to one category or the other a specific ethno/social group or a greater or lesser tendency of the Arab city toward “democratic government.” First of all, it is inappropriate to apply modern concepts such as democracy to the preindustrial world. In addition, after seventy years of fruitless comparison, it is clear that any distinction made between the city founded on precise geometric tissues and the city that develops more organically is purely exploitative. If by “spontaneous” one intends some kind of anarchical settlement, then the “spontaneous” pre-industrial city never existed. As history unfolds, the urban organism always responds to the principle of maximum yield and thereby to a formative logic; even when it avoids the geometry of the grid it does not trust itself to luck but follows typical principles of adherence to the morphology of the place and the cultural instances of society. The crystalline form of an original city is to be found only in illustrated treatises. In the reality of the historical process, the inhabitants of a city quickly see to the deformation of the plan with individual acts of appropriation, as soon as there are fewer guidelines for its design. One controversy in Oriental studies concerns the presumed inferiority of the Arab city - and by extension the Islamic city - because of the weakness of its communal institutions, due to a lack of understanding of the spontaneous fabric.80 This opinion is shared by the frenzied modern tourist and the nineteenth-century traveler on the Grand Tour put off by the maze of alleys, intentionally disorienting routes, and mechanisms used to separate and remove parts of the Arab city from the rest. The traveler unconsciously placed the “chaotic” Arab city in constant competition with the “orderly” classical city, which had been engulfed by it and then consumed from inside. The entire body of Orientalist studies has in the past thirty years undergone a deep revision with a double accusation: on the one hand, an anti-historical attitude for having extended to the entire history of the city in Muslim countries the results of an analysis carried out on traditional Arab and eastern cities during the colonial period, disregarding literary and archaeological data, and on the other hand, superficiality in its attributing the results deduced from Maghrebian cities and in part from Syria to the entire Dar al Islam.81 Cities with unusual founding histories, like the round Baghdad of alMansur - the expression of one man’s power and caprice - are considered by the Orientalists as exceptions. Some scholars associate a labyrinthine and topological morphology with a sociological interpretation according to which the shape of a city evolves without the intervention of the social order that establishes the relationships between members of a community. They describe a culturally sophisticated urban society whose commercial activities are far-flung, but which remains influenced to some degree by tribal models of a nomadic and patriarchal society based on respect among kinsmen.82 Although the
Fig. 167 (Above) Martinafranca, Italy. The morphological diagram shows the presence of closed circular shapes. It is impossible to date their origin, since they could be Arab or pre-Roman structures of the Messapi. The topography could have also contributed to the generation of these closed forms. (Below) Essaouira, Morocco. The Berber quarters of Agadir, Bouaker, and Chabanat. On a flat swampland are shapes similar to those in figure 168 above. Are they traces of a primitive settlement, a form of hitta? The paucity of information only allows hypothesis.
224
after amnesia
Orientalist position has been superceded by this time, uncertainties still exist among scholars regarding the interpretation of the urban form of the Arab city that are its inheritance. The revisionism of the Marçais brothers and other pioneers establishes the fact that the phenomenon is far too complex to be enclosed in a single formal scheme like the one in figure 142. Mediterranean Islam absorbs and synthesizes an assortment of urban traditions, especially classical ones. Absorbed more scientifically than in the West, it is synthesized in original, if unequal, ways in the Maghreb and Mashrek. In the west a strong postfourth-century Berber foundation dominates, while in the east the Persian tradition prevails. In addition, because all great civilizations live on in those that follow them, it is not only important to understand a cultural precedent but also to establish the manner in which it is received and assimilated by its heirs. There is a radical difference between the Levant where, in the 7th century, Islam fit into a tradition of urban growth, and the Maghreb or Spain, where Arab tribes encountered retrogressive societies in proto-urban settlement forms. This duality has precise consequences for urbanism. In Syria, in Palestine and Jordan, Islam guides the Byzantine city to the brink of modernity with no sensible breaks in continuity. In the West, until the thirteenth century, the pattern was to alternate the establishment of cities with regression and a return to nomadic patterns. Keeping in mind the continuity of Oriental urbanism we can debunk another cornerstone of Orientalist thought. According to von Grunebaum, the Arab city tended quickly to corrupt orthogonal patterns,83 but recent archaeological research demonstrates how the deformations in the geometrical warps of the Greco-Roman city in Syria already existed in the third or fourth century, and how other occlusions - columned streets and densified building tissues - can be considered a result of growth associated with an economic boom rather than as a crisis.84 From this point of view there is much less difference between the medieval Italian city and the traditional Arab city beginning in the 14th century than has been described.85 If anything, Arab society carefully managed the transformation of the city by extensively utilizing the principle of maximum yield, carefully metabolizing each piece of the building tissue until the beginning of the 20th century, and taking full advantage of it before introducing a new variation. As a result of this process of gradual change, the Arab city is falsely thought of as immutable. The Orientalist polemic is now obsolete. In spite of the fact that the method of interpretation and our interest are applied above all to the traditional city, which has arrived at the threshold of the contemporary era, it is necessary to reconstruct the founding characteristics and the iconography of the whole Mediterranean city of the south and its evolutionary lines, from the hegira to the 19th century. Jean-Claude Garcin, softening the Orientalist polemic, proposes a suggestive and convincing hypothesis, giving three temporal phases to the Islamic city.86 The first settlements seem to have occurred in a society strongly marked by a tribal structure. Kufa, Basra, Fustat, Kairawan and Baghdad, but in some ways Aleppo and Damascus also, are confederated
Fig. 168 Hypothetical reconstruction of the densification of a primitive semi-nomadic galassia following a multi-centre system.
the urban organism cities with settlements dispersed and kept together by the strong polarity of the Jami Masjid. Immediately this pure tribal structure was challenged by the central power of the Caliph that tends to model the great metropolises of time in the image of the Throne. The radiocentric plan of Baghdad at the time of al Mansur-as reconstructed by Creswell-with its bazaars that from the gates converge at the palace situated in the center of the illustration and the adjacent Jami Masjid, reduced rather to the role of a court chapel, are a metaphor of the Throne that presides over the center of the empire.87 The operation of the monumentalization of the great Mosque of Damascus and the reconstruction of the mosque of Medina in 707 is exemplary for its recovery of all the imperial Byzantine symbology, from the triumphal arch to the processional nave to the tower/minaret. With the crisis of the Abbasids and the disintegration of the eastern Caliphate we have the second phase, and the City of Knights corresponds to the emergence of the military aristocracies, ethnically different from the old dominant Arab classes. The most paradigmatic example in the East is Cairo during the Mameluk era, with its duality of new citadel and indigenous medina, and the squares for military exercises;88 the best known examples in the West are Hafsid Tunis, Merinide Fez and Nasrid Granada. The third phase coincides with the crisis of the 14th century, when the traditional city takes form.89 It corresponds to a time of economic and political crisis of Islam marked by fragile power and reduced financial resources. The built-up area is shut within the derbs and makes the tissues dense, and internal closings multiply. Although each reality is different depending on the place and time, on the basis of principles of typological processuality we know that the traces and forms of the tribal and then caliphate city had to go through the City of Knights and that all the urbanistic operations of the military dynasties must still be present in the mosaic of the traditional city. To the picture furnished by Garcin we should add a note: at the table of the Arab-Mediterranean city there is an uninvited but cumbersome companion: nomad culture. Orientalists insist that the “disorder” of the Arab-Islamic city can be attributed to a return to Bedouin cultural tendencies. This point, indebted mainly to the anti-nomadic philosophy of Ibn Khaldun, has been too readily accepted as self-evident. While the mechanism for fastening the bivouac amsar to the ground and the consolidation of the camp from tents to permanent structures certainly comprise one component that influenced Arab town-planning, their contribution and especially geographical limitations can be better explained. The nomadic camp is usually rounded in form and composed of several compact rows of crouching camels that protect the herd and herdsmen at its center. The polycentric model of many proto-urban settlements is explained by the autonomy of the nomadic camp. When these settlements subsequently came together syncretically, the resulting city could not completely hide its origin; it remained evident in the formal autonomy of the nuclei.90 Similar rounded “nomadic” forms can be found in the Chabanat neighbourhood in Essaouira, but because they developed on flat land they cannot be the result of topography91 (fig. 167). Though it is impossible to draw conclusions based on just a few samples, it would be
225
Fig. 169 Reconstruction of the founding plan of Kufa following Heichem Djait (above) and L. Massignon (below) at the same scale.
226
after amnesia
useful to study both the recently excavated ancient archeological sites in Arabia and Yemen that had no contacts with the classical world and the persistent nomadic customs in various regions of the Arab Mediterranean in Syria or the Sinai. In the case of North Africa it is necessary to take into account the various Berber models of settlements of peasant origin. The villages in the Kabylia and Aurès zones of Algeria, the morphology of which directly reflects topography, have such a low level of hierarchization - neither market nor mosque - that, according to Marçais’s guidelines, they do not qualify as Islamic. The hierarchical structure is reproduced, however, in the territory where the usual ridge-top settlements look to the site of the weekly marketplace as a point of reference and exchange. The latter is located in the center of gravity of the system or between two systems in the bottom of a valley.92 Another type of compact settlement is found on the plateaus behind the rocky Atlas Mountains and in the oasis. The ksar is regular in plan and formed by a single main route with orthogonal residential streets branching off it like fish bones. For reasons of climate, and in order better to utilize the routes as well as the spaces of a house, the former are almost always completely covered. A similar design was imported from the desert by Essaouira in the so-called Qasbah in the 18th century (fig. 158). Because both layouts show no indication of centrality, we may deduce that centrality is culturally acquired and characteristic of an advanced level of development.93 The chronological picture outlined by Garcin functions therefore if we can see in minute detail the continuous, at times strong, presence of the tribal tradition. The model of the tamsir was never obliterated after the 8th century, living on in both the topological plan of both spontaneous and foundation cities, if only at the level of social topography and dislocation of the main functions. Any study of nomadic settlements must examine the ancient khitta tradition: the set of principles established at the time of the Hegira which govern the granting of land to the tribe as a whole.94 This literature is the primary source for studying the urban history of the early Arab-Islamic city, for it documents ancient customs, the influence of which persisted through the Middle Ages. Maqrizi’s text is the first relating to the tamsir of Fustat.95 It describes buildings by categories but unfortunately leaves out topography and the overall layout of the settlement. Not even the reconstructions by Casanova96 and by Sylvie Denoix97 succeed in restoring to it a standard plan, which probably never even existed. Maqrizi explicitly states that the khitta would have frozen the military camp of Fustat along the lines of occupation during a siege, and that this would explain the odd placement of the Jami Masjid and the common spaces that are pushed outwards to the Nile riverbank. The sources describe a similar situation in the founding of Basra, Iraq, where an anarchical occupation of the land seems to have won out over the plans of the Caliph.98Kufa is a key case, though interpretations of its urban history and its sources differ. The two reconstructions to date were made by L. Massignon99 and more recently by Hichem Djait.100 Massignon describes a radiocentric layout in which tribal settlements are
the urban organism scattered around a central site, with roads leading to the main territorial poles - Damascus and Hira for example - branching out from it. The circular city is surrounded by a canal. The second reconstruction describes a centrifugal and standard layout consisting of a sahn, an excavated square area for public use (480 m. x 480 m.), and manahidj, or strips of land 750 m. long, orthogonal to the sides of the sahn and used by the clans for their settlements. The sacred character of the center is emphasized by a ditch that surrounds it, pomoerium-like, and by two isolated enclosures, the Jami Masjid and the Dar al-Umara, that rise from it. The widths of the manahidj vary: the Hamdan clan’s, for example, is 93 m; the Tamim’s is 208 m. They are separated by straight streets measuring 40 cubits, or 21.6 meters in section. Each manahidj is in turn served by main streets measuring 30 cubits in section, which branch out into secondary routes of 20 cubits and residential dead-end alleys of 7 cubits. A djabbana, or open space, is located at the center of each tribal quarter.101 It should be noted that Kufa, like other Tamsir cities, has no walls. The example of Kufa confirms that the Arab-Islamic city is not at the mercy of spontaneous forces acting from below, but also responds to an esprit de geometrie from above. The very practice of the takhfit, akin to the marking out of Kufa’s central square using the Sasanian ritual of shooting arrows, proves that Arabs had long been familiar with concepts like the sacred enclosure, the temple, and founding rites, all essential conditions for the auspicious beginnings of a city.102 From the two reconstructions of Kufa we can infer in any case that the khitta is an a-priori project of agreed-upon city planning, with which the tribe takes possession of the land and at the same time establishes the distances of respect from the other tribes. In the military camp, al-Hira, the city of tents, the khittas of the tribes stand out with their relative areas of expansion and the cemeteries, residential tribal lots in collective possession distant from one another, structured within by a road system of a tree-shaped hierarchy and the strips of separation between the tribes afterwards transformed into actual arteries for the quick movement of troops. All the territorial roads converge in the center where the jami masjid and the dar al Umara, the government palace, are located, just as the streets of the Roman castrum came together in the praetorium. In the tamsir, the city of tents that becomes a city of stone, the relationship between urban network and quarter is ambiguous: the tribal quarter enjoys a strong autonomy, with its own mosque, cemetery and later on, bath, so that we could speak of a particularized urbanistic concept if the central organs of the new foundation were not functionally and physically distinct in a higher hierarchical order. Over time, the strips of respect were invaded by bazaars and by the expansion of the quarters, determining that coincidence of territorial route and market and the hierarchical ambiguity of the foundation routes, of which we spoke in the preceding chapter.103 An important feature of this model is the branching and hierarchized streets. A polycentric settlement in which the parts are separated is a feature of the nomadic-origin model. Figure 168 is an attempt to illustrate the transformation of a polycentric camp/village. The individual closed, rounded forms progressively join with
227
228
after amnesia
routes that have more linear and standard tissues. In opposition to the model of the spontaneous traditional city, scholars have proposed examples of foundation cities without, however, placing the project of the plan in its relative historical context, and forgetting that geometry, while answering to a desire for order that comes from above, has various motivations and objectives that influence its iconography. Above all, the Hellenistic-Roman-Byzantine tradition is the primary reference for the Umayyad caliphate, recognizable in both the so-called Desert Castles and at Anjar.104 The former seem to be the synthesis of two classical models: the castrum and the peristyle; large walled courtyard palaces with bilaterally symmetrical plans as at Mshatta or Ukhaidir, in which the great central square distributes bayts along the edge, that are patio houses. The latter is a plan with two colonnaded streets that intersect at 90 degrees.
Fig. 170 Two cities of Knights compared: Abbasid Samarra and Fatimid Al-Qaira ( From P. Guichard. Les villes d’al Andalus, in P. Cressier and M. Garcia-Arenal, op.cit., pag 44-45).
the urban organism The second tradition is the Persian-Mesopotamian, which is revived with the Abbasid caliphate: on the one hand, the round plan of Baghdad (from 768) that has illustrious precedents in Persia - Creswell lists eleven cases of round or oval cities including Ctesyphon, Darabgird and Sasanide Firuzbad105 - on the other hand Samarra, which develops as a linear city with autonomous quarters along the Tigris, all of which are laid out on a geometric grid.106 Al-Rafiqa near Raqqa (from 770) and Sabra al-Mansuriyya (from 948) near Kairawan refer to the case of Baghdad. Referring instead to the case of Samarra is al-Qahira, the new Fatimide city-palace designed on a geometric grid of 100 black cubits.107 The third tradition derives from the city of knights. We note the tendency of each new dynasty to establish a city next to the previous one, almost as if it wanted to break with history. This tendency was first observed by George Marçais.108 Francisco Benet describes it as a parallax attitude.109 Some of the distinguishing morphological characteristics and planning techniques of an Arab traditional settlement necessary for building a model
229
Fig. 171 Diagram of an Arab city of the Maghreb according to Marçais. Marçais’s idea of the city was based on the centrality of the jami masjid and the antipolarity of the citadel, the cemeteries, and polluting activities. The main bazaars were considered generative elements of urban circulation.
230
after amnesia
that is not completely different from that described by Marçais111 are: 1.Very strict zoning enforcing a clear separation between residential, commercial, and industrial sectors in much the same way as ethnic neighborhoods were separated from one another. The ultimate public space, the suq, is teeming by day, only to empty out completely by sundown. 2. The compartmentalization of the urban fabric in a series of closed and fairly separate nuclei. In periods of growth, residential building modules approach one another without forming a centralizing axis. The suq reveals a similar tendency. Its multitude of doorways, hinges still in place, shows how, in the precolonial period, the space was inhabited and used as a sequence of completely separable segments, like the watertight chambers of a submarine.112 3. A branching pattern of routes and spaces. The distribution of foot traffic creates a highly hierarchized route pattern, similar to the distribution of water by an urban aqueduct web, in which the shift from public to semi-public to private is differentiated by the road section. The main public routes connecting the gates with the central bazaar are more or less continuous in order to facilitate a smooth flow of goods and people and converge toward the central function of the congregational mosque. Secondary public routes act as links to primary routes. The street that in North Africa is called a derb (a term that means both the street and the building tissue that faces it) is an internal passage whose tortured route reflects the building conflicts over ownership of its residents (fig. 109). Finally, the soqaq, which provides access to the innermost houses of a residential block, represents the extreme end of the system. These three characteristics and their variations are so common as to constitute a low level of typological specificity for characteristics that has not escaped urban historians. Returning to the Pauty article, we conclude that the contrast between a planned and a spontaneous city is a contrived one. Cities are both planned and spontaneous, combinations of standard geometries and irregular parts, dependent on cultural forces and the political and economic moment, like a palimpsest which alternates layers of printing and script. Imagine a small first-century BCE Roman military camp, roughly 500 feet in length; radial routes into the territory leave from its three gates, as at Timgad in Algeria (see fig. 121) or Umm Rasas in Jordan.113 Over time, linear hamlets spring up on these routes and soon a fabric with a new system of routes is established. In the interstitial spaces between this fabric and the walls of the camp a third, rather irregular fabric forms. It is made up of synchronic variant building types. This sequence is fairly common. If the area then undergoes a wave of town-planning in the second century CE, the new grid might possibly be rotated with respect to the original because of the dominance of one particular territorial route, and the result will be a new crossed fabric with radial routes branching out from its gates. A new linear fabric with grand imperial-era structures such as a circus, an amphitheatre, a naumachia and an odeum forms along these routes. The spaces between the two fabrics are usually resolved by an infill tissue. Without reviewing the full history of this hypothetical city - it is sufficient to remember the examples of Roman Aleppo described previously - we can see the results
the urban organism of two planning stages fairly close in time. They did not simply generate two grid fabrics, but at least six different forms of fabric, or, two others with planned routes on the suburbs and two infills. Finally we note that the Arab city which is not formally planned is nevertheless always part geometry and part topology. It is topological because it gathers together places, and geometrical or cosmic at the same time because the churches which point eastward and the mosques facing the direction of Mecca often contradict the predominant layout of the urban fabric.114 In conclusion, the difficulty of circumscribing the concept of city is overcome by the definition of urban organism as self-conscious human settlement. Like any other organism, the city is made up of homogeneous and collaborating parts called the urban nucleus, the expression of a growth phase and at times of a different civilization. The urban nucleus is composed of urban tissues that in turn contain building types.Various types of conjunctions between parts have been examined in relation to the compatibility of building types and typical tissues. We have shown how, at the planned grids, linear and encroachment tissues always alternate. The process of self-consciousness of the city during growth is accompanied by a hierarchical growth and nodality for which the intersecting points of the most significant routes of the organism increase in quality, containing special buildings, points of visual, aesthetic and social reference. The growth process of the nodes is continuous and tends to generate ever more complex and specialized nodes up to the so-called poles, or nodes, but it is also dialectical since it is contrasted to anti-nodal spaces, the centre against the outskirts. What is more, over time we can witness an overturning, so that an anti-pole becomes a pole, as in the case of the nineteenth-century railroad station. Up to this point, the urban organism is seen examined closely in its separate components: building type, routes, urban tissues, urban nucleus, nodes and special buildings. The organism must now be re-examined in its unity beginning with the largest scale according to six interpretive schemes: 1.Reading of the territorial plan; 2.Overall reading of the forma urbis; 3.Distinction between the homogeneous parts; 4.Singling out of the natural base of the site (this scheme can also be in the first place); 5.Identification of hierarchies; 6.Identification of pre-existing conditions through the geometry of signs. This last scheme, also called archaeology without excavation, or extensive archaeology, allows us to define fairly accurately the city of the preceding layers. It is a practice invented by the great French historian and archaeologist Jean Sauvaget and perfected by the school of Saverio Muratori. The technique of interpreting signs in the present tissues can signal the circumference of the walls with the placement of gates, the presence of great empty spaces like piazzas and agoras, theatres and the great atopic buildings of the classical world, and so on. Finally, the application of the concept of organism to the Islamic city has shown the deceptiveness of the Orientalist tradition, which saw in that the sign of the crisis of a civilisation incapable of imposing order on its growth. The concept of historical process has demonstrated the artificiality of the contrast between planned city and spontaneous city.
231
232
after amnesia
notes 1. Gustavo Giovannoni was the first to use the biological metaphor, which then passsed directly into the school of Muratori. For this reason all literature on urban morphology remains ambiguous on the subject of the boundaries or levels of urban scale. Often, building tissue and urban fabric are used as interchangeable terms under the generic rubric of urban fabric. It is important to distinguish between these terms, because they represent differences in both scale and quality. The ambiguity between them is also amplified by exceptional cases when a single building tissue, a neighborhood, can be a completely independent organism. Examples of such building tissues can be found in small linear settlements, as on a promontory or a suburb along a mid-level contour, or in an Arab rabat, itself a suburb of an early layout, or a new tissue outside a city gate, which still retains a rural character. It would be more correct to term such building tissues “elementary settlements” since they are primarily composed of dwellings and thus are different from urban fabric. 2. Mukadasi in the Ahsan al-Takasim fi Marifat al-Akalim enumerates as amsars Samarkand, Iranshar, Sharistan, Irbil, Alahwal, Shiraz, Sirdgwan, Mansura, Zubaid, Baghdad, Mecca, Mawsil, Damascus, Fustat, Kairawan and Kurtuba, defined by him as “a balad over which the supreme sultan reigns…” and he lists seventy-seven kasabats. A. Mez in Die Renaissance des Islam (Heidelberg, 1922) mistakenly interpreting Mukadasi, adds two types, al-Nawa= large village and al-Kura=village. In the testimony of classic Arab geographers like Yakut it would appear that the criteria for defining the city from the urbanistic point of view are limited to the presence of the mosque and the market (to which Mukadasi adds the baths). In the formulation of the Hanafite school of law “only in a truly important city must one recite the Friday prayer.” But is it sufficient to give a village a congregational mosque to convert it to a city? And what type does the new city belong to, the medina or the misr? According to Shafite theory the congregational mosque can not be a definition criterion of the status of the city (see E. Herzfeld, “Studies in Architecture,” in Ars Islamica, XIII, [1948] p. 120), as much, as is obvious, as the quality and density of the public buildings. See U. Monnerret de Villard, Introduzione allo studio dell’archeologia islamica, (Florence, 1966),
pp. 92 ff. P. Cuneo, Storia dell’urbanistica. Il mondo islamico (Bari, 1986) addresses this topic extensively, especially in the chapter “Ideologie, teorie, interpretazioni,” pp. 75-82. Instead, a recent monumental essay neglects form and structure to focus on functional aspects: P. Wheatley, The Places where Men Pray Together. Cities in the Islamic Lands, 7th to 10th Centuries (Chicago, 2001). 3. C. Norberg-Schulz, Intentions in Architecture (Cambridge, Mass., 1963), pp. 173-179. 4. This discourse is valid in general terms, and in tissues with a low level of nodality since the introverted space seems the ideal solution for all functional problems of the Arab city. See R. Berardi, “Signification du plan ancien de la ville arabe,” in La Ville Arabe dans l’Islam, op.cit., pp. 180 ff. 5. H. Fathy, Natural Energy and Vernacular Architecture (Chicago, 1986), p. 62. 6. Such slowness and resistance are surprising when one considers how, in the postwar period, migration alone has caused rapid and irreversible change in the structure of all medinas. It would not be too radical to say that in the past fifty years Islamic societies have undergone traumatic changes leading to unsustainable pressures and adverse effects on the medina’s tissue. The old tissue is threatened by the changes in the structure of the traditional family and the solidarity of social groups. Large-scale migration from rural areas has brought people into the core of the medina who did not bring with them an urban way of life. Casablanca exemplifies these phenomena. The periphery of the medina in contact with the colonial city - modern Casablanca - is now a bazaar for the whole metropolis. This new role has determined a typological change from courtyard houses to apartments, which are more easily adapted to modern commercial uses. At the same time, the internal core of the medina, unable to adopt to modern demands because of its limited typological variations, has been abandoned to the poor immigrants and has rapidly degenerated into slums. See A. Adam, “La Prolétarisation de l’habitat dans l’ancienne medina de Casablanca,” Bulletin Economique et Social du Maroc 12 (1949-1950), n. 45, pp. 247-256 and n. 46, pp. 44-50. 7. In Roman Algiers, Icosium, the cardo is the piedmont between Bab Azzoun and Bab el Oued; the decumanus is represented
by rue de la Marine. The two roads are two arcs of a circle, perpendicular only at the graft, where the Sayyida mosque rose. Rue de la Marine forms an arc that follows the curvature of the coast and hence all the foundation roads are not parallel, but circle progressively. R.V. Loliva, La riforma tipologica e costruttiva della casa a corte mediterranea.Tipo edilizio, organismo murario e modello strutturale reagente a sisma nel caso studio di Algeri, doctoral thesis in Architectural Design for the Countries of the Mediterranean, advisor: A. Petruccioli (Bari Polytechnic, 2005). 8. L. Beritic, Utvrdesija Grada Dubrovnika (Zagreb, 1955). 9. J.-C. David, “La Formation du Tissue de la ville arabo-islamique,” p. 155. 10. On the city of Sfax only the article by Michel Van der Meerschen, “La Medina de Sfax: Enquête préliminaire à sa régeneration,” Monumentum n. 8, pp. 6-23. Caniggia’s work on Florence is extremely important for the methodology it employs. See G. L. Maffei, La casa Fiorentina nella storia della citta dalle origini all’Ottocento, introduction by Gianfranco Caniggia (Venice, 1990). 11. G. Downy, A History of Antioch in Syria from Seleucus to the Arab Conquest (Princeton, 1961). 12. J. Weulersse, “Antioche. Essai de géographie urbaine,” in Bulletin d’Etudes Orientales, IV, (1934), pp. 47-48 and drawing V. See also H. Kennedy, “Antioch: from Byzantium to Islam and Back Again,” in J. Rich, ed., The City in Late Antiquity (London, 1992), pp. 181- 198. 13. K. Fisekci, M. Laperchia, P. Martiradonna, P. Patrono, S. Stoppelli, Antakya: dalla forma urbana della città storica al tessuto commerciale degli han, senior thesis, Bari Polytechnic, 2004, reader: A. Petruccioli. 14. S. M. Salsh, Monastir, éssai d’histoire sociale du XIXth siècle (Tunis, 1978). The models differ in morphology but present the same low level of hierarchization as in Kabylia. The graphic analysis of Damascus that I call a “chromatic table” clearly illustrates problems that scholars such as Jean Sauvaget and Dorothee Sack were unaware of or felt were unimportant. The Jami’ Masjid is clearly rotated from the Hellenistic grid based on strigae, but it is parallel to the
233
decumanus called via recta and parallel to the wrap that replaced the old agora. This rotation is not based solely on the direction of the qibla, since it also involves the walls of the temenos of the ancient temple of Jupiter. We can deduce that there are two preexisting grids; the classical monument is not aligned to the Hellenistic grid, but perhaps it is to the earlier one. 15. J. Passini, “Parcellaire et espace urbain médiéval: les villes du chemin de SaintJacques de Compostelle,” in P. Merlin, ed., Morphologie Urbaine et Parcellaire, (Paris, 1988), pp. 187-206. 16. These dimensions correspond to 47.2 x 124 meters and are the same as those already found by E. Wirth (in H. Gaube and E. Wirth, Aleppo. Historische und geographische Beitrage zur baulische Gestaltung, zur sozialen Organisation und zur wirtschaftlichen Dynamik einer vorderasiatischen Fernhandelsmetropole, Wiesbaden, 1984, p. 121, and they are different from the dimensions related by J. Sauvaget (in J. Sauvaget, “Le plan de Laodicée-sur-Mer,” in Bulletin d’Etudes Orientales, IV, Damas, (1934) pp. 81- 114) which were 46 x 120 meters. 17. For these Roman centuriations, see G. A. Neglia, “Persistences and Changes in the Urban Fabric of the Old City of Aleppo,” in Environmental Design, 1-2, (2000-2001), edited by A. Petruccioli, pp. 32- 41, and the doctoral thesis of G. A. Neglia, Città del mediterraneo: Aleppo. Forme e tipi della città intra moenia, Research Doctorate in Architectural Design for the Countries of the Mediterranean, Bari Polytechnic,( 2003 ), advisor: A. Petruccioli. 18. Around Jerusalem there exists a land subdivision on square modules of a heredium. Starting from the castrum intramoenia with its surface of 5 x 4 heredia, on the outside to the west we find an analogous subdivision that, beyond the orographic gradient extends to the facing plateau. This was all part of a division of a centuria, or 4 heredia. Studies on the orientations of the orthogonal fabrics of Jerusalem were carried out beginning in 1988 by this author and A. G. Neglia in the Architecture School of the Bari Polytechnic. See A. Altini, D. Campobasso, E. Masi, G. Mazzone, R. Sardone and M. Simone. Gerusalemme. Analisi morfologica del nucleo antico di Gerusalemme, senior thesis, Bari Polytechnic, ( 2006), advisor: A. Petruccioli.
19. An interesting case is the city of the Abbasid caliph Mutawakill in Samarra: in the tongue of land between the Tigris River and the Al-Qatul canal between two and three kilometres wide and fifteen kilometres long, a monumental median street goes through several quarters planned on a grid and rotated in order to maintain their orthogonality with respect to the Tigris. See E. Herzfeld, Geschichte der Stadt Samarra,VI (Berlin, 1948); The city is composed of three regular gridded fabrics, the first of which is rotated 42 degrees north, the second 12 degrees and the third 20 degrees. J. M. Rogers, “Samarra: A Study in Medieval Town Planning,” in A. H. Hourani and S. M. Stern, eds., The Islamic City (Oxford, 1970), pp. 119- 156, contains a careful examination of the literature on Samarra. We understand the dialectical relationship between tissues of different scales, which is not only one of containment, like Chinese boxes, where we should consider that the urban fabric receives a further typological qualification from the character of building tissues that it contains. In Florence, for example, the Santa Croce quarter is an urban fabric of Gothic plot tissues, while Tripoli (Libya) is an urban fabric of modular crossed tissues. At Fez el-Djedid the urban fabric consists of a topological building tissue made from clusters or cul-de-sacs.
of gravity of the plan. Eugenio Galdieri, during the restoration of the Friday mosque at Isfahan, discovered under the pavement of the prayer hall a masonry cube with a hole for such a stake. See E. Galdieri, Isfahan: Masjid-i Guma (Rome, 1972). This system of tracing domes is also common in Central Asia, and given its simplicity and logic, it may also have been extensively used in the Middle East. 26. On the Roman castrum and the Polybian model, see L. Crema, L’Architettura Romana (Turin, 1959), p. 33.
21. P. Maretto. Realtà naturale e realtà costruita ( Firenze, 1980) p. 229.
27. Of several references, three of the best are Walter Christaller’s Die Zentralen Orte in Süddeutschland (Jena, 1933); English trans., Central Places in Southern Germany (Englewood Cliffs. N.J., 1966); A. Losch, The Economics of Location (New Haven, Conn., 1954); and W. Isard, Methods of Regional Analyses (New York, 1960), have attempted a theoretical and non-empirical solution to the two basic problems: nodes in space, and their hierarchical organization, with particular reference to economics. Due to the abstract format of these studies, for instance, the uniform distribution of the resources and the centers in an isotropic surface, the reading remains incomplete and without corroboration in reality. Furthermore, Christaller’s and Losch’s theories are static in character and unable to accommodate the historical process. However, the continuous contribution of Christaller’s first formulations in 1933 constitutes a precious theoretical body. In particular, I am referring to the concepts of threshold of population, radius of influence of a service and the K factor.
22. M. Serres, La Communication (Paris, 1969).
28. K. Lynch, The Image of the City (Cambridge, Mass., 1960), p. 49.
23. M. Serres, Ibid.; from the introduction.
29. C. Levi-Strauss, Structural Anthropology (New York, 1963), chapter 8.
20. For Naples, instead of innumerable articles, I recommend a direct study of the aerial photographs; see Atlante di Napoli: La forma del centro storico (Venice, 1992).
24. A. Petruccioli, “La moschea,” in G. Strappa, ed., Edilizia per il culto. Chiese, moschee, sinagoghe. Strutture cimiteriali (Turin, 2005), pp. 227- 236. 25. Polarity is characteristic of all centralplan buildings, due to the rotation around a central vertical axis. The structural and virtual axes also converge on this central point. This general principle of polarity is especially pronounced in building consisting of domes constructed without scaffolding. In the simplest version, such building systems involve tracing arches with string from a stake driven into the center
30. S. Tamari, “Aspetti principali dell’urbanesimo musulmano,” in Palladio, I-IV (1966) pp. 45- 82. 31. See the catalogue of the exhibition Sevilla Almohade, curated by M.V. Piechotta and A. Tahiri, Seville, Universidad de Sevilla, (1999) p. 63. 32. This is what happened for example in the Khalig al-Misri canal that ran along the western wall of the Fatimide city of Cairo and acted as a ditch (antinodal and dividing), while after the expansion of
234
after amnesia
the city toward the Nile in the Ottoman period, it was crossed by several bridges and then filled in to become in 1898 the present Shari Port Said (nodal and centralizing). 33. It is more than a mere restoration. This transformation was made by the Italian architect Armando Brasini; in it, the entire composition is centered on the nodal position in the piano nobile of the office of governor Italo Balbo. See C. Conforti, “Armando Brasini’s Architecture at Tripoli,” in Amate Sponde, ed. A. Petruccioli, special issue of Environmental Design 1-2 (1990). 34. A. Petruccioli, “Nodality and Polarity in the Colonial Urban Fabric of Algiers,” in The European House in the Islamic Countries, ed. A. Petruccioli, special issue of Environmental Design, 1-2 (1994). 35. P. Jervis, “Kairouan,” in Urban Fabric, ed. A. Petruccioli, Environmental Design, 1-2 (1989), pp. 36-53 and M. Sakli, “Kairouan,” in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval, Ecole française (Rome, 2000) pp. 57- 85. 36. It is worthwhile to repeat the story: When Saint Francis died, since he did not consider his body worthy of honor or of burial, he asked that it be thrown into the rubbish. After fifty years, and with changes in the rules of the Franciscan order and the creation of a cult of St. Francis, a famous basilica was built by the prior of the Order Fra’ Elia on this earlier neglected dump. 37. E. Celebi, Seyatname, IX, Istanbul (1935) p. 377. 38. M. Cerasi, with E. Bugatti and S. D’Agostino, The Istanbul Divanyolu. A Case Study in Ottoman Urban and Architectural Mentality and Processes of Change, Orient Institut-Ergon Verlag (Würzburg, 2004). 39. S. Muratori, R. Bollati, S. Bollati, and G. Marinucci, Studi per una operante Storia Urbana di Roma (Rome, 1963). 40. On the growth of Cordoba, see E. Levi-Provencal, L’Espagne musulmane au X siècle, pp. 196- 197 and 210- 221; M. Acien Almansa and A.Vallejo Triano, “Cordoue,” in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval, Ecole française ( Rome, 2000) pp. 117- 134. On the Friday Mosque, see M. Nieto Cumplido and C. L. De Tena y Alvear, La mezquita de Cordoba: planos y debujos (Cordoba, 1992).
41. Even if we do not have a precise dating for Sfax and Hammamet in Tunisia, the successive building doubling of the respective jami masjid is visible in the stone elevation. 42. See the recent C. Necipoglu, The Age of Sinan. Architectural Culture in the Ottoman Empire (London, 2005). 43. A. Cutrone, P. Genchi, G.M. Lozupone, L. Massarelli, A. Moccia, D. Panaro, Progetto innovativo della struttura residenziale nella periferia della città di Essaouira, Marocco, senior thesis, Architecture School, Bari Polytechnic (2004), advisor: A. Petruccioli. 44. In reality only the Shafi ’ school of jurisprudence limits each city to one, but even in instances when the Shafi ’ law was enforced, as in Cairo under Salah al-Din, the rule was rarely applied. 45. In a neighborhood built early on, in the high Middle Ages, a parish located at the back of the pertinence strip of the matrix route, since the need of the service defers that of the settlement, covers about a hundred houses, and every parish is about fifty meters from another. But building intensification brings with it a repositioning: the intermediate buildings disappear or remain as oratories, while others become specialized, and are turned into seats of congregations and convents; the parishes at the end of the Middle Ages are at a distance of 200 meters, but their size becomes greater to respond to their new role. 46. J. S. Woodford, The City of Tunis. Evolution of an Urban System (Wisbech, 1990), pp. 64- 84. See also A. Daulatli, Tunis sous les Hafsides: évolution urbaine et activité architecturale (Tunis, 1975). 47. Fatimid Cairo, as reconstructed, is characterized by the strong polarity of the palace, the centre of gravity of the city, with special tissue on both sides of the main urban thoroughfare that acted as a counterpoint to the serial tissue of the residential quarter. Already under the Mamluks, there was a decrease in this palatial tissue, which was in part replaced by special religious and commercial buildings and in part appropriated as residential quarters. New nodal buildings began to appear, such as the two mosques of alHakim and Sultan al-Mu’ayyad, strategically located inside the main gates, and the madrasa-mosque of al-Azhar located in
the centre. At such times the serial tissue, however, remained the dominant element. The modern tissue of Cairo stems from the period of Ottoman domination and is strongly specialized and hierarchized with the extension of the commercial activity beyond the Qasaba. An example of a highly specialized tissue in Cairo is Charaibi Street, studied by the Atelier du Caire under the direction of Philippe Panerai and Sawson Noweir. See J.C. Depaule and S. Noweir, “Atelier du Caire 1. Rue Charaibi,” Bulletin d’informations architecturales, 80 (Nov. 1983). 48.Very instructive is the plan of the ground floors of the walled city in the attached illustration in M. H. Bourgoyne, Mamluk Jerusalem. An Architectural Study (London, 1987). 49. See P. G. Rowe, Making a Middle Landscape (Cambridge, 1991), where he describes areas with low density that form autonomous settlements but cannot be defined as towns. In many Arab countries in the Gulf in the seventies separate compounds for foreigners were built; besides housing they included all the necessary facilities such as a supermarket and even a gymnasium. These settlements were surrounded by a high enclosure so that the Western way of life prohibited to the local population could go on unseen. Ironically, however, this settlement model was rapidly adopted by the local bourgeois society. 50. As Kevin Lynch has pointed out, marginal American settlements undifferentiated in use and type are not memorable; see K. Lynch, The Image of the City (Cambridge, 1960), pp. 123 ff. In European peripheries, on the contrary, it is not the variation of the building type (the apartment house is universally recognized as a building type), but the arbitrary aggregation of the types and the idiosyncratic language of the designers that lead to the illegibility of the tissues. In the 1970s in Italy, Law 167 for LowCost Housing disseminated the urban peripheries of neighbourhoods, similar in concept to urban fabrics, but of improbable design. The Casilino quarter by Ludovico Quaroni in Rome is a prime example; in the absence of an adherence to the place and a continuity with the fabric of the city, every neighbourhood depended on the inventiveness of the designer, and the result was a collage of abstract geometries. 51. E. Lévi-Provencal, La fondation de Fes
235
notes
(Paris, 1939).
pp. 349 ff.
52. J. Tricart, Corso di geografia umana, vol. 2, Habitat urbano (Milan, 1998), pp. 111- 118.
61. In the chromatic table the warps of the second and third group appear the same. Only a precise analysis of the architecture in relation to the slopes reveals origins that clearly belong to different phases.
53. P. Lavedan, Qu’est-ce que l’urbanisme? Introduction à l’histoire de l’urbanisme (Paris, 1926), pp. 91- 105. 54. L. Piccinato, Urbanistica medievale (Florence, 1943); see also M. Morini, Atlante di storia dell’urbanistica, dalla preistoria all’inizio del XX secolo (Milan, 1963).
62. C. Rowe and F. Koetter, Collage City (Milan, 1981). 63. M. Poete, op. cit.
55. M. R. G. Conzen, Alnwick, Northumberland: A Study in Town Plan Analysis, Institute of British Geographers, 27 (London, 1960).
64. P. Lavedan, op. cit.; see also M. R. G. Conzen. “The Use of Town Plans in the Study of Urban History,” in H. J. Dynos, ed., The Study of Urban History (London, 1968).
56. J. W. Reps, The Making of Urban America. A History of City Planning in the United States (Princeton, 1965).
65. J. Navarro Palazòn, Una casa islamica en Murcia. Estudio de su ajuar (siglo XIII) (Murcia, 1991).
57. S. Muratori, Civiltà e territorio ( Roma, 1967).
66. M. Sartre, Bostra dès origins à l’Islam (Paris, 1985).
58. There is a margin of error intrinsic to cadastral plans and architectural surveys because measurements are taken by hand, but these can be ignored, as they remain constant throughout. Similarly, in tracing parallel alignments we can also ignore minimum variations within the systems that stem from minor discrepancies in building reconstruction, such as the consolidation of the borders of a plot. Aside from a case where occasional signs do not repeat, it is advisable to concentrate on homogeneous warps and the longest segments.
67. André Bazzana argues in his book Maisons d’Al-Andalous, chap. 3, n. 11, in favor of so-called extensive archaeology, a method of reading the built landscape with extensive use of aerophotos and cartography. This method used by the entire French archaeological school based in Spain in the Casa de Velasquez at Madrid is concerned with a non-intrusive analysis. In Italy the Scuola di Topografia Antica operates in Rome. See P. Sommella, “Finalità e metodi della lettura storica in centri a continuità di vita,” in Archeologia Medievali VI (1979), pp. 105- 128, which works according to the principle of the continuity of settlement.
59. The use of chromatic drawings has the advantage of showing the various warps clearly. No one has clearly delineated the grid that covers the whole of Fatimid Cairo including the few exceptions, which could represent primitive villages. The width of a module is 28 meters (84 feet), which might be a Greek measure. 60. The Greek implants are a variant of this group. The tissue has a prevailing direction because it is influenced by the Greek land division system - per strigas et scamma according to which only the transverse dimension is fixed, while the other is indefinite. See G. Schmiedt, “Contributo della fotografia aerea alla ricostruzione del paesaggio rurale nell’Alto Mediolevo,” in XIII Settimana di Studio del Centro Italiano Studi Alto Medioevo (Spoleto, 1966), pp. 773 ff.; G. Schmiedt and R. Chevallier, “Caulonia e Metaponto,” in L’Universo 39,
68. P. Pinon, “La transformaciòn desde la ciudad antique a la ciudad medieval: permanencia y transformaciòn de los tejidos urbanos en el mediterràneo oriental,” in La ciudad medieval: de la casa al tejido urbano,” Universidad de CastillaLa Mancha (Cuenca, 2001). On Muslim Spain in general, see: B. Pavon, Ciudades hispanomusulmanas, Mapfre (Madrid, 1992), in particular the chapter “Ciudades hispanomusulmanas asentadas sobre las antiguas,” pp. 185- 296. 69. The sanctuary of Ba’al in Palmyra consists of a monumental enclosure with a single monumental access to the west, a portico on four sides and a cella at the center of a vast empty esplanade. In the Ayyubid era the whole sanctuary
was transformed into a fortress with the blocking up of the only gate and the erection of a wall along the perimeter. In time, the interior was jammed with the inhabitants of the oasis until it formed a compact Arab tissue and the cella was transformed into a mosque. Only in 1930 was the complex liberated by the authorities of the French Mandate. Aerial photographs of the time show, in addition to the routes and openings, a rather regular tissue owing to the geometric plan of the sacred quadrilateral, with some straight streets including the bazaar and a summary warp of rectangular modules. The horseshoe shaped tissue of the suq has a higher density. See T. Wiegand, Palmyra, Ergebnisse der Expeditionen von 1902 und 1917 (Berlin, 1932). There is an image of the total encroachment and two photographs from within of some small streets of the city built in the sanctuary; E. Will, Les Palmyriens, La Venis de sables (Ier siècle avant-IIIème siècle après J.-C., (Paris, 1992), shows the sanctuary of Ba’al after the first partial removal of the buildings installed in the enclosure of the sanctuary itself. 70. J. Sauvaget, “Le plan antique de Damas,” Syria 26 (1949), p. 332. N. Elisséeff, “Damas à la lumière des theories de Jean Sauvaget,” in A. H. Hourani and S. M. Stern, eds., The Islamic City (Oxford, 1970), pp. 157178. For a different interpretation see: E. Will, “Damas antique,” in Syria, LXXI, pp. 1- 43. For a synthesis, see T. Bianquis, “Damas,” in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval (Rome, 2000), pp. 37- 55. 71. Three routes, of which one aligned with the decumanus axis of the colony, point toward the eastern gate of the medieval city built in Byzantine times, over an existing ancient gate. Two routes - one an extension toward the gate of Mardin of the decumanus axis of the ancient city - converge at the Byzantine church of Saint Stephen converted later on into a congregational mosque, placing the ancient east gate of the colony. Z. Akkoyunlu, Geleneksel Urfa evlerinin mimari ozellikleri (Ankara, 1989). 72. Every classical town had at least one theatre. Philadelphia/Amman and Gerasa even had two.Very soon these theaters were not in use, since the performances were prohibited by the Christians. The theatres in each town went through radical transformations; for example, Bosra in Syria and Arles in Provence were turned into fortresses in the Ayyubid era when the
236
after amnesia
town was consolidated. 73. Following Jean Sauvaget,ibid., pp. 346349. In the area one can see a bent alley named zqaq es-saha, street of the square. 74. Examples of total obstruction can be seen at Latakia. See Sauvaget’s plan of Latakia in correspondence with the tetrapylon, in “Le plan de Laodicée-surMer,” op.cit. 75. For historic information on Hammamet see L. Micara, “Il rilievo architettonico della medina di Hammamet (Tunisia),” in Storia della Città, 46 (1988), pp. 81- 84. 76. The Roman linear and surface measures are: the foot = 29.57 cm; the passus = 5 Roman feet = 1.48 meters; the pertica = 10 passus = 14.8 meters; the actus = 120 Roman feet = 35.4 meters and the mile = 5,000 Roman feet are the urban and territorial measures. The surface measures start from the actus = the square of 120 feet x 120 feet, then the jugerum, equivalent to 2 actus of the land that can be worked by 2 oxes in one day; the heredium = 4 actus, that is 240 feet x 240 feet, and finally the centuria, which is the best known and first territorial module of 2,400 x 2,400 feet. 77. For a comparison, the Byzantine measures are: the foot (pous) of Hagia Sofia = 31.23 cm; the foot dimension in Byzantine Empire varies from 30.8 cm to 32 cm. It is derived from the Greek foot (31.6 cm) and not the Roman foot. See the terms “Metrology,” “Pous,” “Daktilos” in the Oxford Dictionary of Byzantium, ed. A. Kazhdan, 3 vols. (Oxford, 1991); P. Underwood, “Some Principles of Measure in the Architecture of Justinian,” in Cahiers Archéologique 3 (1948), pp. 6474. For local measures, see E. Schilbad, Bizantinische Métrologie (Thessaloniki, 1982); T. Thieme. Le dessin d’architecture dans les societés antiques (Leiden, 1985), pp. 291-308, reports two other units: the daktilos = 1.95 cm; the spithame (palm) = 23.4 cm, also called basilike or imperial. Multiples of these measures are: the orgyia = 6 feet = 96 daktyloi or 1.87 cms. The standard land measure in use was a larger orgyia equal to 108 daktiloi. 78. For the Islamic world, we are unable to reconstruct the system of measures, but two things are clear. First, the measures changed from town to town, just as in medieval Europe. Second, the system of measure
was unlike the Roman system. According to Walter Hinz there were two leading measures of the cubit: the standard dira = 49.875 cm, that divided by 24 gives the inch = asba w= 2.078 cm and the black dira = 54.04 cm with the relative asba = 2.252 cms. See W. Hinz, “Islamische Masse und Gewichte,” in Handbuch der Orientalistik, 1, 1 (Leiden, 1955), p. 54; K. A. C. Creswell has also deduced the black dira to have been the basis of Cairo’s Nilometer; see Early Muslim Architecture (Oxford, 1940), pp. 290. In the Abbasid period the black cubit spread all over the Mediterranean; we find it in Spain and the Maghreb in the time of Caliph Ma’mun (813-833). On the ancient measures we have more diverging data: Omar’s dira would have been 72.815 cm; the dira at Kufa seems to be 50.3 cms. In the medieval Levant every city had its own cubit: Cairo 58.187 cm; Damascus 63.035 cm; Aleppo 67.9 cm; and Jerusalem (as per T. Tobler, Denkblaetter aus Jerusalem [St. Gallen - Konstanz, 1853], pp. 279), 64.77 cm. None of these measures can be compared with the Roman classical ones, except perhaps for Cairo’s black dira, which is almost equivalent to two Roman feet. 79. E. Pauty, “Villes spontanées et villes crées en Islam,” Annales de l’Institut d’Etudes Orientales 9 (1951), pp. 52-75. 80. It is worth summarizing the various Orientalist positions on the Islamic city. The merit of extending the discourse from the monuments to the urban fabric goes to the brothers Georges Marçais and William Marçais who emphasized the essentially urban role of religion. See G. Marçais, “L’urbanisme musulman,” in Mélanges d’histoire et d’archéologie de l’occident musulman, 1: Articles et Conférences de Georges Marçais (Algiers, 1957), pp. 219-231. See also W. Marçais, “L’islamisme et la vie urbaine,” in Comptes rendus de l’Académie des Inscriptions et Belles Lettres (1928), pp. 86-100; Gustave E. von Grunebaum and Robert Brunschvig share Marçais’s spatial model as well as a certain resistance to the “irrationality” of the Arab urban fabric. Their starting points are, however, different. Grunebaum maintains a socio-political position, and insists on the lack of urban and collective institutions, little democratic political life, and animosity between quarters; see G. E. von Grunebaum, “Islam: Essays in the Nature and Growth of a Cultural Tradition”, special issue of The American Anthropologist, no. 81 (1955), in particular, chapter 8, “The Structure of the Muslim Town,” pp. 141-158. Brunschvig’s
bias stems from the analysis of the legal systems, and notes that the structure of the Islamic cities is based on customary laws. See R. Brunschvig, “Urbanisme médiéval et droit musulman,” Revue des Etudes Islamiques 15 (1947), pp. 127-155. Claude Caen notes how guilds are another element of disaggregation, because their religious orientation is towards mysticism or the Shi’a. See C. Caen, “Mouvements populaires et autonomisme urbain dans l’Asie musulmane du moyen age,” Arabica 5 (1958), pp. 225-250, and 6 (1959), pp. 25-56, 233-260. The French geographer Xavier de Planhol disregards the Islamic cities with unusual sharpness. The scarce urban sense, the inability to form a coherent physical and social structure is attributed to the nomadic origins of the Arabs. Planhol sees the nomadic element as persistent and continously challenging society, which itself is seen as a unitarian body. The city is, to use André Miquel’s phrase, “the tribe that has settled.” See X. de Planhol, Les Fondements géographiques de l’histoire de l’Islam (Paris, 1968). Part of the text is summarized in idem, The World of Islam (Ithaca, 1959). Ernst Egli denies the Islamic city any specificity and design intentionality. E. Egli, Geschichte des Staedtebaues, 3 vols., Das Mittelalter (Erlenbach, 1962) pp. 266 ff. Only after the sixties, with the conference at Oxford, did these Orientalist propositions begin to be challenged. In this conference, Albert Hourani noted the African origin of the Marçais model, and every other participant underlined the impossibility of generalizing from a single model since the Islamic city is a complex phenomenon, extended in both geography and in time (see A. Hourani and S. M. Stern, eds., The Islamic City: A Colloquium [Oxford, 1970]). The German geographer Eugen Wirth, after a careful comparative analysis of functions and morphologies of Arab-Islamic cities, came to the conclusion that the codified characteristics of the Islamic city - arboreal system of routes, culs-de-sac, bent entrances, and courtyard type - are also present in Mesopotamian cities like Ur and Kirkuk. Wirth therefore proposed substituting the “Oriental City” for the term Islamic city. 81. A coherent critique was made by J. Abu-Lughod, “The Islamic City. Historic Myth, Islamic essence and Contemporary Relevance,” in International Journal of Middle East Studies, 19/2, (1987), pp. 155-176. 82. Jacques Berque, who belongs to the
237
notes
French sociological school, was the first to praise the spaces of the Arab traditional city and contrast them with the lack of quality in a modern city such as Casablanca. But Berque exalted only the civitas, regarding the Arab Islamic city as the site of commercial exchange and religious testimony, in contrast to the Marçais brothers, for whom the monuments and the material components of the Islamic city provided its formative character. 83. G. von Grunebaum, “Structure of the Muslim Town,” op. cit. 84. This hypothesis, which I support, was synthesized by Hugh Kennedy in “From Polis to Medina: urban change in late antique and early Islamic Syria,” in Past & Present, 106 (1985) pp. 3-27, and in “The Towns of Bilad al-Sham and the Arab Conquest,” in M. A. Bakhit and M. Asfourm eds., Proceedings of the Symposium on Bilad alSham during the Byzantine Period” (Amman, 1983), vol. 2, pp. 88-99. The encroachments of the shops in the Umayyad and Abbasid eras in the colonnaded street of Palmyra should be read in this sense. See K. Al Asad and F. M. Stepniowski, “The Umayad Suq in Palmyra,” in Damaszener Mitteilungen, 4 (1989), pp. 205-211. 85. L. Torrés-Balbas, “Les villes musulmanes d’Espagne et leur urbanisation,” in Annales de l’Institut d’Etudes Orientales, 7 (1947), p. 12. D. Whitcomb proposes a revision of the Orientalist thesis showing how diverse new settlements of the Umayyad era, built behind the Byzantine cities, were misrs with a regular plan as at Tiberias and al-Ramlah, similar to that of Anjar. D. Whitcomb, “The misr of Ayla,” in T. E. Levy, ed., “The Archaeology of Society in the Holy Land,” London (1995) pp. 155-170. 86. J. C. Garcin, “Le Caire et l’évolution urbaine des pays musulmans, “ in Annales islamologiques, 25 (1991) pp. 289-304. 87. K. A. C. Creswell, Early Muslim Architecture: Ummayyads, early Abbasids & Tulunids (Oxford, 1932-40). See also J. Lassner, “Notes on the Topography of Baghdad: the Systematic Descriptions of the City and the Khatib al-Baghdadi,” in Journal of the American Oriental Society, LXIII, pp. 458-469, and the more recent F. Micheau, “Baghdad,” in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval (Rome, 2000), pp. 87116.
88. N. D. Mackenzie, Ayyubid Cairo. A Topographical Study (Cairo, 1992). 89. If it were possible to establish a date for the beginning of this long crisis it would be reasonable to place it in 1348, when the terrible Black Plague massacred the cities of the Mediterranean. 90. “At Kufah the inhabitants of one quarter required a guide when they entered another,” Creswell, op.cit., p. 30. 91. At Martinafranca in southern Italy, which was under Arab domination for more than a century, I was able to isolate the same polycentric layout of antagonistic systems. In spite of the topolines, I suspect that this is a similar phenomenon. 92. A. Petruccioli, “Strutture fortificate dell’Atlante algerino,” in I sistemi difensivi del bacino del Mediterraneo (Crotone, 1992). 93. In the Kabyl villages of the Beni Yenni tribe, the mosque for instance appeared only in the 1980s and as a political decision. It is not located in the fabric of the village but at the entrance, and abutting a paved road; it will be a node for the future growth of the village. 94. Khitta is the concession of land by the caliph to a clan or tribe; it was used in the first and second century of the Hegira. The aim was to settle the nomadic tribes, give form to the Umma, and break or at least limit the bonds of kinship in favor of communal solidarity. Khitta is a legal instrument that permits the passage from the hira, the semi-nomadic encampment, to the tamsir, the fixed military encampments that would later become a city. See S.Tamari, op.cit., p. 75. Another form of concession was the dar, an individual concession reserved for the companion of the Prophet and his clients. Ibn Duqmaq, in describing the tamsir of Fustat, also mentions the dar with the names of individuals and Ya’kubi in the Kitab al Baldan uses it in the same sense. At Fez, founded in 800 by the Idrisses, the first urban space was the qirtas, made up of the Arab and Berber tribes. Idriss II established the qirtas, “giving to the Qaysite Arabs from the Gate of Ifriqiya to the Gate of Iron in the quarter of the Qairawin, and established the tribe of Azd next to them and the Yasubs next to these. The Berber tribes of Sinhaga, Lunata, Masmuda and al-Sayhan each in its place ordering them to work and cultivate the land.” I. Azi Zar, Rawd
al-qirtas, I, (Valencia, 1964), pp. 89-90. For Fez in general see H. Ferhat, “Fes,” in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval (Rome, 2000), pp. 215-233. 95. All the original sources on Fustat have been lost, but fortunately they are summarized in the monumental work of al-Maqrizi, Al-Marwa’ir wa’l-i’tibar fi khiata Misr wa al-Kahira. 96. P. Casanova, “Essai de reconstitution topografique de la ville de al-Foustat ou Misr, “ in Mémoires IFAO, 35, Cairo (1919). 97. S. Denoix, op.cit. Chapter 4, note 36. 98. S. Ahmad al-Ali, “Khilat al-Basra,” in Sumer 8 (1952), pp. 72 ff. 99. L. Massignon. “Explication du plan de Kufa” in Mélanges Maspero (Cairo, 19351940), vol.3. 100. H. Djait, Al-Kufa: Naissance de la ville islamique (Paris, 1986). 101. We know that 150 years later, when Abu Mikhriaf began writing, this Cartesian implant was already in crisis. The reconstruction imagined by Hichem Djait (which I have simplified in figure 169) is too rigid. Such a plan would not have been able to absorb the expansion of the tribes and the newcomers. This explains how the expansion of the quarters occurred by occupying the few open spaces, that is, the public routes and the sahn. 102. U. Monneret de Villard, op.cit., Ch.VI. 103. Gayraud states, “We must note that this irregular urbanism presents a network of streets of uncertain width and plan. And that the street is not an axis of circulation… but rather a space left free among different constructions; there is, therefore, a preeminence of the latter over the former.” A. Fuad ad Sayyed and R. Gayraud, “Fustat-Le Caire à l’époque fatimide, “ in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval (Rome, 2000), p. 143. 104. On the so-called Desert Castles, see O. Grabar, “Umayyad Palaces and the Abbasid Revolution,” in Studia Islamica, XVIII, (1963), pp. 15-18. On Anjar, see M. Chehab, “The Umayyad Palace at Anjar,” in Ars Orientalis,V (1963), pp. 17-25. 105. K. A. C. Creswell, Early Muslim
238
after amnesia
Architecture, op.cit, pp. 18-22. 106. A. Northedge, Samarra: Residenz der Abbasidenkalifen, pp. 836-892 (Tubingen, 1990), speaks of disjunctive appearance with regard to the plan of the city because of the rotations of the quarters, determined by the orientation of the palaces situated at the ends of the streets that compose the grids. 107. The study of the alignments and modules of Al-Qahira, conducted on the present land register at a 1:1000 scale, in the course “Typological Analysis” that I taught at M.I.T. in the spring of 1996, revealed that the city intra-muros insists on a land division with a 54-meter module (equivalent to 100 black cubits in Cairo) and with a submodule one half that, confirming the planned character of the Fatimid capital. D. BehrensAbouseif, S. Denoix and J.-C. Garcin, “Le Caire,” in J.-C. Garcin, ed., Grandes villes méditerranéennes du monde musulman médiéval (Rome, 2000), pp. 177-203. Saladin, authorizing the demolition of the Fatimid palaces of al-Qahira, transfers the center of power into his citadel and encourages the transformation of old Cairo into a religious and commercial center. 108. G. Marçais, “L’urbanisme musulman,” op.cit., p. 221. 109. F. Benet, “The Ideology of Islamic Urbanization,” in B. Anderson, ed., Urbanism and Urbanization (Leiden, 1963), pp. 111-126. The article focuses on the strategies the caliphs used to settle the nomads and reinforce the unity of the Umma. 110. As noted by E. Wirth, “Die Regelhafte Raumorganization des Almohadischen Planconzepts,” Madrider Mitteilungen 34 (1993), pp. 348-368. 111. In addition to the plan in which I summed up Marçais’s idea, other authors have attempted diagrams: M. de Epalza, “Espacios y sus funciones en la ciudad arabe,” in Simposio Internacional sobre la ciudad islamica (Zaragoza, 1991), pp. 26-27. 112. R. Berardi, “Alla ricerca di un alfabeto urbano: la medina di Tunisi,” in Necropoli 9-10 (1970), p. 27. S. Bianca, Urban Form in the Arab World (New York, 2000), pp. 154157. 113. S. Tateo, Il paesaggio del Limes Arabicus: dai castra difensivi romani ai palazzi Omayyadi,
doctoral thesis in Architectural Design in the Countries of the Mediterranean (2003), pp. 254-264, advisor: A. Petruccioli. 114. M. Bonine, “The Sacred Direction and City Structure: a Preliminary Analysis of the Islamic Cities of Morocco,” in Muqarnas, 7, (1990) pp. 50-72. We recall that in India a particular case of cosmic orientation is often linked to the direction in space of the zodiac sign of the founder, according to a tradition adopted there by the least Orthodox Muslim dynasties. See A. Petruccioli, “Ad quadratum. Notes on Deccani Town Planning,” in A. Dallapiccola and S. Zingel-Avé Lallemant, eds., Islam and Indian Regions (Heidelberg, 1993), pp. 193-202.